{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'powersocket', 'check', 'television', 'wine', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table', 'clothes_skirt']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": enter office, find computer, turn it on,,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TOUCH] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'freezer', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'button', 'bookmark', 'ceilinglamp', 'window', 'couch', 'microphone', 'pillow', 'wall', 'pasta', 'tape', 'box', 'novel', 'curtain', 'doorjamb', 'hairbrush', 'door', 'home_office', 'table', 'bathroom_cabinet', 'soap', 'tooth_paste', 'clothes_shirt']\nTask:Text friends while sitting on couch\nresponse:\n", "output": {"thoughts": "Text while sitting", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[TOUCH] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'couch', 'candle', 'clothes_hat', 'pillow', 'wall', 'oven', 'coffe_maker', 'television', 'curtain', 'diary', 'towel_rack', 'mouse', 'dishrack', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_pizza', 'juice', 'crayon', 'blow_dryer', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'mat', 'tvstand', 'faucet', 'wall', 'powersocket', 'bench', 'tape', 'photoframe', 'clothes_gloves', 'food_donut', 'after_shave', 'doorjamb', 'dishrack', 'door', 'sink', 'bathroom_cabinet', 'soap']\nTask:Clean sink\nresponse:\n", "output": {"thoughts": "Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.", "actions": ["[WALK] <bathroom>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[WIPE] <faucet>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'scissors', 'crayon', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'bookmark', 'toaster', 'ceilinglamp', 'stovefan', 'faucet', 'window', 'homework', 'couch', 'wall', 'child', 'bench', 'photoframe', 'newspaper', 'coffe_maker', 'razor', 'television', 'kitchen_counter', 'cupboard', 'table', 'coffee']\nTask:Let baby learn how to walk\nresponse:\n{\n    \"thoughts\": The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[GREET] <child>', '[FIND] <couch>', '[TURNTO] <couch>', '[POINTAT] <couch>', '[TURNTO] <child>', '[LOOKAT] <child>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'ceiling', 'character', 'toaster', 'ceilinglamp', 'tvstand', 'stereo', 'deck_of_cards', 'oil', 'wall', 'bench', 'powersocket', 'novel', 'coffe_maker', 'food_food', 'television', 'tray', 'rag', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'scrabble']\nTask:Make popcorn\nresponse:\n", "output": {"thoughts": "Lay popcorn bag flat inside microwave and press popcorn button to start.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <microwave>", "[TURNTO] <microwave>", "[OPEN] <microwave>", "[PUTIN] <food_food> <microwave>", "[CLOSE] <microwave>", "[SWITCHON] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'detergent', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'bathroom_counter', 'phone', 'folder', 'man', 'keyboard', 'mop_bucket', 'tvstand', 'faucet', 'picture', 'plate', 'mousepad', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'check', 'food_onion', 'clothes_scarf', 'spoon', 'light', 'dishwasher']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": When someone knocks I n the door walk through the hallway and greet them.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'walllamp', 'food_butter', 'toothbrush', 'knifeblock', 'floor', 'character', 'bookshelf', 'food_cake', 'mat', 'shoe-shine_kit', 'shower', 'mousepad', 'couch', 'microphone', 'cup', 'newspaper', 'mouthwash', 'television', 'coffee_filter', 'food_donut', 'console', 'kitchen_counter', 'headset', 'dresser', 'cupboard', 'clothes_dress', 'home_office']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Wait until doorbell rings, look to see who it is outside and open the door if you feel so inclined.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'sponge', 'scissors', 'chair', 'dog', 'drawing', 'bills', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'ceilinglamp', 'shoe-shine_kit', 'window', 'foundation', 'couch', 'wall', 'bench', 'basket_for_clothes', 'bed', 'novel', 'television', 'kitchen_counter', 'dresser', 'cupboard', 'bathroom_cabinet', 'light', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Grab book.  Read it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <dresser>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'microwave', 'jelly', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'ceiling', 'character', 'table_cloth', 'toaster', 'stovefan', 'mat', 'iron', 'pillow', 'dining_room', 'wall', 'powersocket', 'cup', 'photoframe', 'coffe_maker', 'television', 'pencil', 'kitchen_counter', 'hairbrush', 'cupboard', 'door', 'dishwasher', 'clothes_skirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'beer', 'microwave', 'bathroom', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'toaster', 'brush', 'plate', 'shower', 'clothes_socks', 'dining_room', 'wall', 'powersocket', 'tape', 'box', 'cup', 'coffe_maker', 'food_food', 'doorjamb', 'centerpiece', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['beer', 'walllamp', 'drawing', 'orchid', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'couch', 'candle', 'pillow', 'wall', 'nightstand', 'powersocket', 'keys', 'television', 'curtain', 'doorjamb', 'door', 'sink', 'table']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sit down on couch in front of TV.  Use remote to turn on the TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'ground_coffee', 'walllamp', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'character', 'computer', 'piano_bench', 'stovefan', 'toaster', 'tvstand', 'faucet', 'oil', 'pillow', 'wall', 'oven', 'powersocket', 'food_food', 'water', 'tray', 'after_shave', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Get my coffee pot going,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffee_pot>', '[FIND] <coffee_pot>', '[GRAB] <coffee_pot>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffee_pot>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffee_pot>', '[GRAB] <ground_coffee>', '[PUTOBJBACK] <ground_coffee>', '[WALK] <oven>', '[PUTBACK] <coffee_pot> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'jelly', 'freezer', 'walllamp', 'ground_coffee', 'knife', 'knifeblock', 'cpuscreen', 'orchid', 'needle', 'floor', 'ceiling', 'desk', 'stovefan', 'ceilinglamp', 'faucet', 'mousepad', 'clothes_hat', 'wall', 'oven', 'cup', 'food_food', 'toothbrush_holder', 'tray', 'rag', 'kitchen_counter', 'cupboard', 'sink', 'bag']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Take out a rag. Turn on faucet and get water on the rag. Clean counter off.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <rag> <sink>", "[SWITCHOFF] <faucet>", "[GRAB] <rag>", "[SQUEEZE] <rag>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'food_salt', 'freezer', 'drawing', 'knife', 'cpuscreen', 'knifeblock', 'floor', 'character', 'bookshelf', 'laptop', 'plate', 'mousepad', 'candle', 'dining_room', 'wall', 'light_bulb', 'food_food', 'coffe_maker', 'tray', 'kitchen_counter', 'chessboard', 'dishrack', 'shoes', 'cupboard', 'laundry_detergent', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'walllamp', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'face', 'computer', 'bookmark', 'mat', 'faucet', 'mousepad', 'clothes_hat', 'towel', 'dining_room', 'wall', 'wall_clock', 'nightstand', 'face_soap', 'basket_for_clothes', 'newspaper', 'tray', 'rag', 'mouse', 'dishrack', 'sink', 'bathroom_cabinet']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <rag>", "[GRAB] <rag>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[WASH] <rag>", "[FIND] <face_soap>", "[GRAB] <face_soap>", "[POUR] <face_soap> <rag>", "[PUTOBJBACK] <face_soap>", "[FIND] <face>", "[SCRUB] <face>", "[PUTBACK] <rag> <sink>", "[RINSE] <face>", "[RINSE] <rag>", "[SWITCHOFF] <faucet>", "[GRAB] <rag>", "[PUTOBJBACK] <rag>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <face>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'detergent', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'folder', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'tvstand', 'shower', 'homework', 'oil', 'dustpan', 'dining_room', 'wall', 'powersocket', 'bench', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into the bedroom and walk to the light switch. Flip the switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'tablelamp', 'microwave', 'dog', 'walllamp', 'trashcan', 'knifeblock', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'faucet', 'ceilinglamp', 'oil', 'wall', 'oven', 'coffe_maker', 'drinking_glass', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'bag', 'dishwasher', 'colander']\nTask:Unload dishwasher\nresponse:\n", "output": {"thoughts": "Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[PUTBACK] <drinking_glass> <cupboard>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[PUTIN] <fork> <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTIN] <knife> <kitchen_cabinet>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTIN] <spoon> <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTIN] <plate> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'walllamp', 'drawing', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'stovefan', 'plate', 'faucet', 'tvstand', 'window', 'dish_soap', 'towel', 'dining_room', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'food_food', 'curtain', 'kitchen_counter', 'cupboard', 'food_steak', 'sink']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walk to Kitchen, find sink, find dishes and dish soap. Put soap on plate. open faucet, rinse dishes close faucet.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <plate>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <plate>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cat', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'food_noodles', 'mousepad', 'wall', 'oven', 'check', 'video_game_controller', 'television', 'oven_mitts', 'after_shave', 'food_onion', 'centerpiece', 'mouse', 'door', 'table', 'coffee']\nTask:Use computer\nresponse:\n", "output": {"thoughts": "Sit at my desk and grade papers on my computer", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'jelly', 'freezer', 'food_carrot', 'drawing', 'floor', 'ceiling', 'desk', 'envelope', 'table_cloth', 'mat', 'faucet', 'mousepad', 'couch', 'pillow', 'towel', 'wall', 'cup', 'food_food', 'mouthwash', 'television', 'water', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'food_steak', 'light', 'dishwasher', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[TURNTO] <cup>', '[LOOKAT] <cup>', '[WALK] <cup>', '[FIND] <freezer>', '[TURNTO] <freezer>', '[LOOKAT] <freezer>', '[WALK] <freezer>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>']\n}\n", "input": "\nScene_Objects:['food_carrot', 'milk', 'dough', 'knifeblock', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'mat', 'faucet', 'shower', 'stereo', 'towel', 'pillow', 'wall', 'bench', 'cup', 'bed', 'photoframe', 'coffe_maker', 'wooden_spoon', 'after_shave', 'rag', 'mouse', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Turn on the water faucet. Wet my hands and turn off the faucet. Put soap on my hands. Rub hands together, making sure the soap covers all areas of the hands. Scrub hands for about 20 seconds. Turn on the water water faucet. Rinse hands under the running water. Turn off the water faucet. Dry hands with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'chair', 'knifeblock', 'hanger', 'floor', 'toaster', 'stovefan', 'shoe-shine_kit', 'faucet', 'homework', 'food_noodles', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'cup', 'newspaper', 'coffee_filter', 'food_rice', 'console', 'mouse', 'kitchen_counter', 'dishrack', 'comb', 'cupboard', 'door', 'sink']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to the kitchen, walk to the cabinet, pic up a glass. carry the glass to the sink. turn on the water and fill up the glass. drink the water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['mop', 'freezer', 'trashcan', 'milk', 'knifeblock', 'orchid', 'clothes_underwear', 'bookshelf', 'purse', 'thread', 'computer', 'ceilinglamp', 'tvstand', 'iron', 'couch', 'microphone', 'wall', 'cutting_board', 'basket_for_clothes', 'keys', 'toothbrush_holder', 'oven_mitts', 'curtain', 'doorjamb', 'diary', 'chessboard', 'home_office', 'bathroom_cabinet', 'clothes_skirt', 'light']\nTask:Hang up car keys\nresponse:\n", "output": {"thoughts": "Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <purse>", "[FIND] <purse>", "[OPEN] <purse>", "[FIND] <keys>", "[GRAB] <keys>", "[CLOSE] <purse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'drawing', 'hanger', 'floor', 'phone', 'ceilinglamp', 'plate', 'iron', 'candle', 'pillow', 'wall', 'bench', 'powersocket', 'oven', 'food_food', 'video_game_controller', 'check', 'wooden_spoon', 'after_shave', 'doorjamb', 'shoes', 'dresser', 'cupboard', 'clothes_dress', 'door', 'conditioner', 'sink', 'table', 'light']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": Put food on the table, sit down, and eat the food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[PUTBACK] <plate> <table>', '[FIND] <chair>', '[SIT] <chair>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'orchid', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'microphone', 'pillow', 'wall', 'wall_clock', 'photoframe', 'television', 'toothbrush_holder', 'food_donut', 'doorjamb', 'shaving_cream', 'mouse', 'door', 'sink', 'clothes_scarf', 'napkin', 'table']\nTask:Straighten pictures on wall\nresponse:\n", "output": {"thoughts": "Move the pictures/paintings until they are straight.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'dry_pasta', 'walllamp', 'trashcan', 'food_carrot', 'dough', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'wall', 'wall_clock', 'oven', 'food_cheese', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher']\nTask:Keep an eye on stove as something is cooking\nresponse:\n{\n    \"thoughts\": Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <pot>', '[TURNTO] <pot>', '[LOOKAT] <pot>', '[FIND] <wall_clock>', '[TURNTO] <wall_clock>', '[LOOKAT] <wall_clock>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'measuring_cup', 'freezer', 'nail_polish', 'food_carrot', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'plate', 'mat', 'mousepad', 'food_peanut_butter', 'pillow', 'wall', 'powersocket', 'bench', 'food_vegetable', 'food_food', 'doorjamb', 'mouse', 'door', 'light', 'table', 'scrabble']\nTask:Fix snack\nresponse:\n", "output": {"thoughts": "Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <kitchen_cabinet>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTBACK] <food_carrot> <plate>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[PUTBACK] <food_vegetable> <plate>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_salt', 'measuring_cup', 'toilet', 'jelly', 'drawing', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'alcohol', 'mat', 'shower', 'foundation', 'electric_shaver', 'mousepad', 'wall', 'toy', 'basket_for_clothes', 'photoframe', 'wine', 'dresser', 'laundry_detergent', 'food_steak', 'door', 'napkin', 'bathroom_cabinet', 'dishwasher']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTBACK] <toy> <dresser>', '[CLOSE] <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sheets', 'pot', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mail', 'tvstand', 'stereo', 'window', 'mousepad', 'clothes_socks', 'wall', 'nightstand', 'powersocket', 'pasta', 'oven', 'light_bulb', 'coffe_maker', 'food_rice', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'door', 'light']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'toilet', 'chair', 'freezer', 'detergent', 'food_carrot', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'bedroom', 'piano_bench', 'ceilinglamp', 'mat', 'shower', 'faucet', 'oil', 'food_orange', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'check', 'doorjamb', 'vacuum_cleaner', 'door', 'table', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Put the groceries on the table into the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTBACK] <food_chicken> <freezer>', '[CLOSE] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_carrot> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'dough', 'drawing', 'orchid', 'needle', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'bookmark', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'food_kiwi', 'wall', 'oven', 'photoframe', 'novel', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office', 'table']\nTask:Admire art\nresponse:\n", "output": {"thoughts": "Looking at painting, art", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'microwave', 'toilet', 'jelly', 'bathroom_counter', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'food_noodles', 'clothes_socks', 'mousepad', 'closetdrawer', 'food_kiwi', 'wall', 'cutting_board', 'cup', 'coffe_maker', 'toothbrush_holder', 'television', 'oven_mitts', 'water', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'soap', 'tooth_paste']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, walk to cupboard, open cupboard, grab cup, walk to sink, pour water in cup, drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <sink>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'dough', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'clothes_jacket', 'homework', 'mousepad', 'towel', 'wall', 'nightstand', 'bed', 'novel', 'coffe_maker', 'toothbrush_holder', 'ice', 'diary', 'kettle', 'mouse', 'door', 'spoon', 'instrument_violin']\nTask:Study\nresponse:\n", "output": {"thoughts": "Do homework or read books while sitting on the chair.", "actions": ["[WALK] <bedroom>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'window', 'mousepad', 'couch', 'wall', 'powersocket', 'bed', 'curtain', 'doorjamb', 'shaving_cream', 'mouse', 'comb', 'door', 'napkin', 'table', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I turn on computer. I use keyboard to type. I open browser. I search.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'video_game_console', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'candle', 'broom', 'wall', 'powersocket', 'check', 'video_game_controller', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'hairbrush', 'door', 'table', 'dishwasher', 'dirt']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'chair', 'bathroom', 'drawing', 'food_egg', 'floor', 'desk', 'bookshelf', 'computer', 'mat', 'stereo', 'shower', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'cup', 'doorjamb', 'mouse', 'kitchen_counter', 'pajamas', 'controller', 'door', 'clothes_scarf', 'light', 'dishwasher', 'board_game']\nTask:Pull up carpet\nresponse:\n{\n    \"thoughts\": Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mat>', '[FIND] <mat>', '[TOUCH] <mat>', '[PULL] <mat>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'band-aids', 'cat', 'food_carrot', 'knife', 'foundation', 'electric_shaver', 'mousepad', 'clothes_socks', 'pillow', 'dining_room', 'nightstand', 'oven', 'toy', 'check', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'chessboard', 'cupboard', 'conditioner', 'door', 'food_steak', 'clothes_skirt', 'duster']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Get a broom or duster with long handle. Dust blades of ceiling fan.", "actions": ["[WALK] <bedroom>", "[WALK] <duster>", "[FIND] <duster>", "[GRAB] <duster>", "[PUTOBJBACK] <duster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'character', 'video_game_console', 'bedroom', 'computer', 'button', 'table_cloth', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'box', 'video_game_controller', 'television', 'curtain', 'pencil', 'dresser', 'door', 'table']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'blender', 'toilet', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'toy', 'cup', 'bed', 'keys', 'television', 'curtain', 'doorjamb', 'shaving_cream', 'door', 'table', 'shoe_rack']\nTask:Watch  TV\nresponse:\n", "output": {"thoughts": "Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'tablelamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'purse', 'computer', 'clothes_jacket', 'toaster', 'stereo', 'mousepad', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'bed', 'novel', 'mouse', 'shoes', 'dresser', 'clothes_dress', 'light', 'table']\nTask:Get ready for day\nresponse:\n{\n    \"thoughts\": Grab shoes, purse etc and head out the door!,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTON] <clothes_dress>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>', '[FIND] <purse>', '[GRAB] <purse>', '[CLOSE] <dresser>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'drawing', 'needle', 'knife', 'clothes_underwear', 'hanger', 'floor', 'envelope', 'computer', 'ceilinglamp', 'oil', 'mousepad', 'pillow', 'bowl', 'wall', 'nightstand', 'child', 'bed', 'novel', 'food_food', 'coffe_maker', 'television', 'doorjamb', 'dishrack', 'comb', 'laundry_detergent', 'conditioner', 'sink', 'instrument_violin']\nTask:Story reading time\nresponse:\n", "output": {"thoughts": "Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[SIT] <bed>", "[TURNTO] <child>", "[LOOKAT] <child>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'envelope', 'computer', 'mop_bucket', 'stovefan', 'mat', 'tvstand', 'toaster', 'foundation', 'mousepad', 'pillow', 'wall', 'standingmirror', 'photoframe', 'toilet_paper', 'oven_mitts', 'tray', 'mouse', 'sink']\nTask:Read book\nresponse:\n{\n    \"thoughts\": go to living room,find the text book on the self ,sat on the chair ,open and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'filing_cabinet', 'microwave', 'walllamp', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'computer', 'piano_bench', 'mat', 'tvstand', 'faucet', 'shower', 'homework', 'clothes_socks', 'broom', 'wall', 'bench', 'cup', 'rag', 'kitchen_counter', 'laundry_detergent', 'door', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'spectacles', 'pot', 'microwave', 'knifeblock', 'floor', 'ceiling', 'folder', 'character', 'toaster', 'plate', 'drying_rack', 'faucet', 'dish_soap', 'couch', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'hairdryer', 'light', 'dirt', 'coffee']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": I walk in to the Kitchen. I find a plate, a sponge, and the dish soap. I find the faucet. I turn the faucet on. I pour water on the sponge and plate. I add dish soap to the sponge. I put the dish soap back. I scrub the plate. I rinse the plate. I place the plate on the drying rack. I rinse the sponge. I turn the faucet off.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <sponge>', '[GRAB] <sponge>', '[RINSE] <sponge>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[WASH] <plate>', '[RINSE] <plate>', '[FIND] <drying_rack>', '[PUTBACK] <plate> <plate>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['glue', 'microwave', 'freezer', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'purse', 'mail', 'ceilinglamp', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'bed', 'light_bulb', 'check', 'food_food', 'curtain', 'kitchen_counter', 'chessboard', 'dishrack', 'dresser', 'laundry_detergent', 'door', 'light', 'shoe_rack']\nTask:Open window\nresponse:\n", "output": {"thoughts": "Unlock the latch at the top. Push the bottom half of the window upwards.", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'dog', 'walllamp', 'clothes_underwear', 'bathroom_counter', 'phone', 'character', 'desk', 'tea_bag', 'envelope', 'toaster', 'ceilinglamp', 'window', 'mousepad', 'food_kiwi', 'nightstand', 'cutting_board', 'photoframe', 'coffe_maker', 'television', 'tray', 'rag', 'doorjamb', 'diary', 'cupboard', 'hairdryer', 'home_office', 'light', 'table', 'creditcard']\nTask:Receive credit card\nresponse:\n{\n    \"thoughts\": hear the dooble, go to door, take de envelope, open it with a scissor, take the creditcard,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <envelope>', '[FIND] <envelope>', '[GRAB] <envelope>', '[WALK] <home_office>', '[WALK] <scissors>', '[FIND] <scissors>', '[OPEN] <envelope>', '[FIND] <creditcard>', '[GRAB] <creditcard>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'walllamp', 'bathroom', 'detergent', 'knifeblock', 'bathroom_counter', 'purse', 'alcohol', 'bookmark', 'stovefan', 'toaster', 'stereo', 'food_jam', 'dining_room', 'powersocket', 'food_food', 'mouthwash', 'tray', 'after_shave', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office', 'laser_pointer']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'trashcan', 'floor', 'tea_bag', 'thread', 'toaster', 'ceilinglamp', 'plate', 'dish_soap', 'wall', 'bed', 'coffe_maker', 'food_food', 'check', 'oven_mitts', 'tray', 'doorjamb', 'pencil', 'kitchen_counter', 'hairbrush', 'cupboard', 'hair', 'controller', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'toilet', 'walllamp', 'orchid', 'hanger', 'floor', 'ceiling', 'shampoo', 'bedroom', 'desk', 'envelope', 'table_cloth', 'bookmark', 'brush', 'toaster', 'window', 'couch', 'dustpan', 'food_jam', 'pillow', 'dining_room', 'wall', 'photoframe', 'check', 'tray', 'curtain', 'lighter', 'dresser', 'light']\nTask:Gaze out window\nresponse:\n", "output": {"thoughts": "Look at the yard from the window", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[FIND] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'toilet', 'freezer', 'wallshelf', 'hanger', 'bathroom_counter', 'bookshelf', 'mail', 'brush', 'mat', 'shoe-shine_kit', 'homework', 'closetdrawer', 'couch', 'food_orange', 'towel', 'dining_room', 'wall', 'powersocket', 'box', 'novel', 'check', 'coffe_maker', 'tray', 'doorjamb', 'diary', 'cupboard', 'home_office', 'bathroom_cabinet', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Pick up the book, open to current page, read the words.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['glue', 'chair', 'bathroom', 'detergent', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'purse', 'face', 'mat', 'tvstand', 'faucet', 'mousepad', 'wall', 'bench', 'face_soap', 'newspaper', 'novel', 'ice', 'curtain', 'doorjamb', 'chessboard', 'dresser', 'conditioner', 'sink', 'bathroom_cabinet', 'light', 'table']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Run warm water. Cleanse my face with soap. Rinse my face with water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <face_soap>", "[GRAB] <face_soap>", "[FIND] <face>", "[SCRUB] <face>", "[PUTOBJBACK] <face_soap>", "[WASH] <face>", "[RINSE] <face>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'walllamp', 'woman', 'drawing', 'phone', 'desk', 'man', 'toaster', 'laptop', 'mat', 'shower', 'electric_shaver', 'oil', 'form', 'pillow', 'dining_room', 'oven', 'shelf', 'bed', 'standingmirror', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'tooth_paste', 'scrabble']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'measuring_cup', 'walllamp', 'detergent', 'cat', 'milk', 'knifeblock', 'floor', 'ceiling', 'character', 'folder', 'bookshelf', 'piano_bench', 'table_cloth', 'faucet', 'window', 'mousepad', 'dining_room', 'wall', 'powersocket', 'food_cheese', 'bed', 'check', 'kitchen_counter', 'cupboard', 'sink', 'clothes_scarf', 'clothes_skirt']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'tablelamp', 'mop', 'chair', 'freezer', 'walllamp', 'dough', 'drawing', 'knifeblock', 'wallshelf', 'orchid', 'floor', 'ceiling', 'mop_bucket', 'toaster', 'stovefan', 'faucet', 'bowl', 'wall', 'oven', 'bench', 'toy', 'tray', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": go to the kitchen and cook some thing to eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <oven>', '[OPEN] <oven>', '[FIND] <food_pizza>', '[GRAB] <food_pizza>', '[PUTIN] <food_pizza> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'scissors', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'faucet', 'mousepad', 'food_orange', 'wall', 'powersocket', 'wooden_spoon', 'after_shave', 'doorjamb', 'shaving_cream', 'mouse', 'door', 'table']\nTask:Computer work\nresponse:\n", "output": {"thoughts": "Find the computer, Turn on the computer, find keyboard, find mouse, go online", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[POINTAT] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'mop', 'toilet', 'cat', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'orchid', 'floor', 'ceiling', 'computer', 'clothes_jacket', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'couch', 'wall', 'nightstand', 'food_food', 'mouthwash', 'television', 'food_donut', 'curtain', 'after_shave', 'dresser', 'table']\nTask:Look at painting\nresponse:\n{\n    \"thoughts\": Walk to the painting.  Look up at the painting.  Admire it on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_pants', 'form', 'pillow', 'towel', 'wall', 'powersocket', 'bench', 'newspaper', 'television', 'curtain', 'doorjamb', 'mouse', 'hair', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'walllamp', 'cat', 'milk', 'orchid', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'bookshelf', 'toaster', 'ceilinglamp', 'mat', 'stereo', 'picture', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'basket_for_clothes', 'photoframe', 'doorjamb', 'door', 'light', 'table']\nTask:Look at painting\nresponse:\n{\n    \"thoughts\": Walk to the painting.  Look up at the painting.  Admire it on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'chair', 'freezer', 'bathroom', 'food_butter', 'drawing', 'food_egg', 'floor', 'computer', 'mat', 'tvstand', 'stereo', 'laptop', 'clothes_pants', 'towel', 'wall', 'nightstand', 'tape', 'bed', 'food_food', 'coffe_maker', 'television', 'oven_mitts', 'tray', 'curtain', 'diary', 'cards', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Turn on radio, put on favorite channel, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'microwave', 'chair', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'computer', 'table_cloth', 'tvstand', 'shower', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'television', 'tray', 'doorjamb', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'board_game']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk into living room, walk to light, turn on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['food_bread', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'face', 'toaster', 'mat', 'faucet', 'foundation', 'mousepad', 'towel', 'dining_room', 'broom', 'wall', 'bowl', 'bench', 'face_soap', 'food_cheese', 'coin', 'water', 'after_shave', 'rag', 'curtain', 'mouse', 'conditioner', 'sink', 'napkin', 'bathroom_cabinet', 'hands_both']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.", "actions": ["[WALK] <bathroom>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <rag> <sink>", "[GRAB] <rag>", "[FIND] <face_soap>", "[GRAB] <face_soap>", "[POUR] <face_soap> <rag>", "[PUTOBJBACK] <face_soap>", "[FIND] <face>", "[SCRUB] <face>", "[PUTBACK] <rag> <sink>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>", "[RINSE] <face>", "[WASH] <rag>", "[GRAB] <rag>", "[PUTOBJBACK] <rag>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <face>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'tablelamp', 'freezer', 'bathroom', 'milk', 'drawing', 'floor', 'ceiling', 'bedroom', 'alcohol', 'computer', 'table_cloth', 'stovefan', 'mat', 'shower', 'oil', 'couch', 'pillow', 'wall', 'cutting_board', 'newspaper', 'food_food', 'television', 'kettle', 'kitchen_counter', 'cupboard', 'clothes_dress', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find groceries. Grab groceries. Find fridge. Walk to fridge. Put groceries in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['food_bread', 'food_ice_cream', 'freezer', 'milk', 'drawing', 'knife', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'dvd_player', 'stamp', 'desk', 'keyboard', 'computer', 'button', 'mat', 'plate', 'ceilinglamp', 'mousepad', 'candle', 'pillow', 'wall', 'bench', 'powersocket', 'cup', 'food_food', 'doorjamb', 'mouse', 'door', 'napkin', 'table', 'light']\nTask:Cut bread\nresponse:\n", "output": {"thoughts": "I get a slice of bread, grab a knife, and then cut the piece of bread in half.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <kitchen_cabinet>", "[FIND] <table>", "[PUTBACK] <plate> <table>", "[PUTBACK] <knife> <table>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[PUTBACK] <food_bread> <plate>", "[GRAB] <knife>", "[CUT] <food_bread>", "[PUTOBJBACK] <knife>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_ice_cream', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'longboard', 'floor', 'bookshelf', 'computer', 'faucet', 'balanceball', 'window', 'iron', 'food_peanut_butter', 'food_jam', 'wall', 'nightstand', 'bench', 'powersocket', 'basket_for_clothes', 'novel', 'check', 'coffee_filter', 'shaving_cream', 'pencil', 'maindoor', 'sink', 'hairdryer']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter the Home office, then walk to the phone, pickup the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'jelly', 'freezer', 'walllamp', 'nail_polish', 'knifeblock', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'button', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'mousepad', 'pillow', 'wall', 'oven', 'powersocket', 'bench', 'cup', 'food_food', 'food_rice', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'table', 'board_game']\nTask:Grab some juice\nresponse:\n", "output": {"thoughts": "I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTOBJBACK] <juice>", "[CLOSE] <freezer>", "[FIND] <table>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_ice_cream', 'toilet', 'walllamp', 'dough', 'drawing', 'knifeblock', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bedroom', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'basket_for_clothes', 'bed', 'food_food', 'television', 'doorjamb', 'mouse', 'pencil', 'dishrack', 'door', 'home_office', 'light', 'bathroom_cabinet']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Grasp light switch, flip light switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[TURNTO] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'crayon', 'tea', 'beer', 'microwave', 'toilet', 'trashcan', 'food_carrot', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'bedroom', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'window', 'dining_room', 'food_kiwi', 'nightstand', 'bed', 'photoframe', 'doorjamb', 'dishrack', 'cupboard', 'sink', 'home_office']\nTask:Open bathroom window\nresponse:\n", "output": {"thoughts": "Make sure bathroom window is unlocked. Lift up bathroom window so it is open.", "actions": ["[WALK] <bathroom>", "[WALK] <window>", "[FIND] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'bench', 'bed', 'coffe_maker', 'television', 'food_donut', 'doorjamb', 'mouse', 'door', 'table', 'shoe_rack']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to living-room. Look at the desk where Computer stays. Sit in chair. Switch on computer. Work on computer. Switch off computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[TURNTO] <desk>', '[LOOKAT] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[WATCH] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'freezer', 'walllamp', 'milk', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'alcohol', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'mousepad', 'wall', 'oven', 'powersocket', 'cup', 'ice', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Get something to drink\nresponse:\n", "output": {"thoughts": "Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[CLOSE] <freezer>", "[POUR] <milk> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'freezer', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'brush', 'tvstand', 'stereo', 'window', 'closetdrawer', 'towel', 'wall', 'powersocket', 'tape', 'food_food', 'television', 'curtain', 'doorjamb', 'vacuum_cleaner', 'dresser', 'door', 'clothes_scarf', 'dishwasher', 'board_game']\nTask:vacuum carpet\nresponse:\n{\n    \"thoughts\": I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'cat', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'folder', 'desk', 'keyboard', 'computer', 'faucet', 'stereo', 'mousepad', 'food_peanut_butter', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'check', 'toothbrush_holder', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'napkin', 'light']\nTask:Read\nresponse:\n", "output": {"thoughts": "I like to read books in my bedroom.", "actions": ["[WALK] <bedroom>", "[WALK] <desk>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'tea', 'chair', 'freezer', 'ground_coffee', 'drawing', 'orchid', 'floor', 'purse', 'faucet', 'balanceball', 'food_noodles', 'food_peanut_butter', 'wall', 'powersocket', 'toy', 'bed', 'food_food', 'check', 'food_donut', 'doorjamb', 'pencil', 'laundry_detergent', 'door', 'bag', 'napkin', 'light', 'table']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Grab toys off floor.  Put it in a container.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'freezer', 'walllamp', 'food_carrot', 'drawing', 'clothes_underwear', 'bathroom_counter', 'floor', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'floor_lamp', 'foundation', 'couch', 'pillow', 'bowl', 'dining_room', 'wall', 'nightstand', 'oven', 'bed', 'food_donut', 'console', 'cupboard', 'bathroom_cabinet']\nTask:Turn on lights\nresponse:\n", "output": {"thoughts": "Walk to your lamp, switch it on, and lay on your bed", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'jelly', 'walllamp', 'bathroom', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'couch', 'food_jam', 'wall', 'novel', 'television', 'curtain', 'doorjamb', 'comb', 'clothes_dress', 'sink', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'food_bread', 'filing_cabinet', 'bathroom', 'ceiling', 'phone', 'keyboard', 'purse', 'bookshelf', 'laptop', 'ceilinglamp', 'faucet', 'nightstand', 'bench', 'photoframe', 'bed', 'clothes_gloves', 'food_food', 'coffe_maker', 'television', 'toothbrush_holder', 'curtain', 'shaving_cream', 'pencil', 'dresser', 'conditioner', 'door', 'spoon', 'bathroom_cabinet', 'table']\nTask:Shut front door\nresponse:\n", "output": {"thoughts": "Close the storm door. Lock the storm door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'mat', 'window', 'clothes_socks', 'couch', 'pillow', 'towel', 'broom', 'wall', 'bench', 'oven', 'toy', 'bed', 'coffe_maker', 'television', 'wooden_spoon', 'curtain', 'rag', 'doorjamb', 'food_onion', 'towel_rack', 'mouse', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Find couch. Sit on couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'food_ice_cream', 'chair', 'walllamp', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'mousepad', 'clothes_hat', 'towel', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'shoes', 'door', 'bathroom_cabinet']\nTask:Put in chair\nresponse:\n", "output": {"thoughts": "Take a comfortable chair. Place it in the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[GRAB] <chair>", "[WALK] <home_office>", "[WALK] <floor>", "[FIND] <floor>", "[PUTBACK] <chair> <floor>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'floor', 'character', 'bookshelf', 'computer', 'toaster', 'mat', 'plate', 'window', 'food_noodles', 'food_orange', 'wall', 'oven', 'photoframe', 'keys', 'bed', 'coffe_maker', 'toothbrush_holder', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'clothes_dress', 'napkin', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sponge', 'band-aids', 'toothbrush', 'orchid', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'faucet', 'picture', 'shower', 'mousepad', 'candle', 'food_orange', 'oven', 'basket_for_clothes', 'bed', 'check', 'food_food', 'television', 'tray', 'after_shave', 'headset', 'comb', 'dresser', 'laundry_detergent', 'napkin', 'bathroom_cabinet']\nTask:Sit in chair\nresponse:\n", "output": {"thoughts": "I sit in a chair to rest and eat.", "actions": ["[WALK] <dining_room>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'microwave', 'freezer', 'detergent', 'drawing', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'ceilinglamp', 'iron', 'faucet', 'window', 'homework', 'bowl', 'wall', 'wall_clock', 'bench', 'shelf', 'newspaper', 'mouthwash', 'food_donut', 'doorjamb', 'coffee_pot', 'door', 'sink', 'bathroom_cabinet', 'light', 'board_game']\nTask:Turn light off\nresponse:\n{\n    \"thoughts\": Put the light switch into the down, or off, position.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'blow_dryer', 'walllamp', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'faucet', 'picture', 'iron', 'mousepad', 'bowl', 'wall', 'powersocket', 'bench', 'photoframe', 'toilet_paper', 'coffe_maker', 'toothbrush_holder', 'tray', 'doorjamb', 'dishrack', 'sink', 'bathroom_cabinet', 'home_office', 'light']\nTask:Get toilet paper\nresponse:\n", "output": {"thoughts": "There are extra toilet papers in the bottom cabinet.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <toilet_paper>", "[GRAB] <toilet_paper>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'trashcan', 'orchid', 'cpuscreen', 'ceiling', 'bedroom', 'table_cloth', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'powersocket', 'bench', 'bed', 'novel', 'food_food', 'mouthwash', 'coffe_maker', 'oven_mitts', 'curtain', 'doorjamb', 'cards', 'dresser', 'cupboard', 'bathtub', 'home_office', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the book shelf, get a book, open the book and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'sheets', 'walllamp', 'knifeblock', 'drawing', 'orchid', 'floor', 'bedroom', 'mail', 'food_bacon', 'laptop', 'mat', 'plate', 'faucet', 'food_orange', 'pillow', 'wall', 'bench', 'box', 'bed', 'photoframe', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'tooth_paste']\nTask:Bring dirty plate to sink\nresponse:\n", "output": {"thoughts": "Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[TURNTO] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[PUTBACK] <plate> <sink>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'ground_coffee', 'cpuscreen', 'floor', 'ceiling', 'keyboard', 'toaster', 'ceilinglamp', 'mat', 'faucet', 'food_orange', 'bowl', 'wall', 'wall_clock', 'cutting_board', 'bed', 'photoframe', 'coffe_maker', 'mouthwash', 'coffee_filter', 'water', 'doorjamb', 'pencil', 'kitchen_counter', 'hair', 'controller', 'sink', 'soap']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[GRAB] <ground_coffee>', '[PUTOBJBACK] <ground_coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_pizza', 'filing_cabinet', 'microwave', 'chair', 'walllamp', 'orchid', 'knifeblock', 'bathroom_counter', 'ceiling', 'phone', 'stamp', 'keyboard', 'thread', 'computer', 'stovefan', 'toaster', 'plate', 'shower', 'mousepad', 'form', 'wall', 'food_cheese', 'bed', 'console', 'headset', 'dresser', 'spoon', 'bathroom_cabinet', 'bag']\nTask:Walk to room\nresponse:\n", "output": {"thoughts": "Leave the living room and enter the hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'freezer', 'bathroom', 'detergent', 'bathroom_counter', 'floor', 'bedroom', 'toaster', 'plate', 'shower', 'dining_room', 'nightstand', 'oven', 'cup', 'coffe_maker', 'food_food', 'television', 'tray', 'kitchen_counter', 'dresser', 'door', 'sink', 'bathtub', 'home_office', 'table', 'dishwasher', 'soap', 'board_game']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['juice', 'beer', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'folder', 'keyboard', 'computer', 'ceilinglamp', 'window', 'mousepad', 'couch', 'pillow', 'bowl', 'food_kiwi', 'wall', 'powersocket', 'food_food', 'curtain', 'doorjamb', 'mouse', 'door', 'light', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a can of beer, open it, sit on couch, relax", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <beer>", "[GRAB] <beer>", "[CLOSE] <freezer>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[DRINK] <beer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'bathroom', 'drawing', 'knife', 'clothes_underwear', 'floor', 'mail', 'food_bacon', 'piano_bench', 'stovefan', 'ceilinglamp', 'plate', 'deck_of_cards', 'toaster', 'oil', 'wall', 'wall_clock', 'oven', 'cup', 'novel', 'food_food', 'television', 'oven_mitts', 'cards', 'cupboard', 'conditioner', 'table']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": Sit down at the dinning table with my family and eat dinner.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[GRAB] <food_food>', '[FIND] <plate>', '[PUTBACK] <food_food> <plate>', '[GRAB] <food_food>', '[PUTOBJBACK] <food_food>', '[FIND] <fork>', '[GRAB] <fork>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['chair', 'ground_coffee', 'bathroom', 'detergent', 'dough', 'knifeblock', 'drawing', 'food_egg', 'floor', 'bedroom', 'keyboard', 'purse', 'mop_bucket', 'stovefan', 'mousepad', 'food_orange', 'bed', 'check', 'toilet_paper', 'food_rice', 'tray', 'shaving_cream', 'mouse', 'cd', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'clothes_skirt', 'creditcard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Twist the lamp switch until the light goes on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'detergent', 'hanger', 'floor', 'phone', 'stamp', 'toaster', 'balanceball', 'oil', 'pillow', 'dining_room', 'wall_clock', 'toy', 'newspaper', 'keys', 'photoframe', 'food_food', 'clothes_gloves', 'rag', 'doorjamb', 'mouse', 'dishrack', 'dresser', 'cupboard', 'hairdryer', 'spoon', 'table', 'light', 'scrabble']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Pet Cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'jelly', 'freezer', 'bathroom', 'trashcan', 'milk', 'toothbrush', 'drawing', 'bathroom_counter', 'floor', 'desk', 'bookshelf', 'alcohol', 'table_cloth', 'faucet', 'shower', 'form', 'powersocket', 'bench', 'bed', 'food_food', 'kitchen_counter', 'headset', 'vacuum_cleaner', 'cupboard', 'door', 'clothes_scarf', 'food_steak', 'soap']\nTask:Grab things\nresponse:\n", "output": {"thoughts": "The robot will need some kind of hands or grabbing device to transport objects.", "actions": ["[WALK] <novel>", "[TURNTO] <novel>", "[GRAB] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'orchid', 'drawing', 'cpuscreen', 'bills', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'mousepad', 'clothes_socks', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'mouse', 'shoes', 'comb', 'door', 'light']\nTask:Check email\nresponse:\n{\n    \"thoughts\": Turn on laptop.  Connect to Internet.  Type in the url.  Check for new email messages.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <mail>', '[GRAB] <mail>', '[READ] <mail>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'food_salt', 'chair', 'walllamp', 'detergent', 'drawing', 'clothes_underwear', 'floor', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'food_noodles', 'garbage_can', 'mousepad', 'microphone', 'wall', 'nightstand', 'food_vegetable', 'bed', 'food_food', 'coffee_filter', 'mouse', 'sink', 'home_office', 'table', 'dishwasher', 'pantry']\nTask:Organize pantry\nresponse:\n", "output": {"thoughts": "Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.", "actions": ["[WALK] <dining_room>", "[WALK] <pantry>", "[FIND] <food_food>", "[GRAB] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[FIND] <garbage_can>", "[PUTBACK] <food_food> <garbage_can>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <dry_pasta>", "[GRAB] <dry_pasta>", "[TURNTO] <dry_pasta>", "[LOOKAT] <dry_pasta>", "[PUTOBJBACK] <dry_pasta>", "[FIND] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[PUSH] <food_food>", "[FIND] <food_noodles>", "[GRAB] <food_noodles>", "[TURNTO] <food_noodles>", "[LOOKAT] <food_noodles>", "[PUTOBJBACK] <food_noodles>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'scissors', 'microwave', 'ceilingfan', 'drawing', 'knifeblock', 'orchid', 'floor', 'ceiling', 'desk', 'bookshelf', 'purse', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'nightstand', 'child', 'cup', 'television', 'curtain', 'centerpiece', 'food_steak', 'hairdryer', 'table', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on tv.  Sit with kids.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[FIND] <child>', '[GREET] <child>', '[FIND] <child>', '[GREET] <child>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'picture', 'homework', 'mousepad', 'wall', 'oven', 'toy', 'newspaper', 'check', 'television', 'mouse', 'laundry_detergent', 'conditioner', 'laser_pointer']\nTask:Push in desk chair\nresponse:\n", "output": {"thoughts": "Push the desk chair until it is under the desk.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[TOUCH] <chair>", "[PUSH] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_salt', 'chair', 'detergent', 'trashcan', 'dough', 'drawing', 'character', 'computer', 'button', 'brush', 'mat', 'couch', 'clothes_pants', 'clothes_hat', 'wall', 'nightstand', 'basket_for_clothes', 'food_cheese', 'rag', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'laundry_detergent', 'conditioner', 'sink', 'home_office', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'juice', 'microwave', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'shower', 'faucet', 'window', 'mousepad', 'food_orange', 'wall', 'television', 'curtain', 'after_shave', 'mouse', 'dresser', 'door', 'napkin', 'table']\nTask:Clean screen\nresponse:\n", "output": {"thoughts": "Use a soft rag to wipe the screen until there are no marks on it.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <computer>", "[WIPE] <computer>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'detergent', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'character', 'stamp', 'keyboard', 'toaster', 'plate', 'faucet', 'tvstand', 'bowl', 'wall', 'oven', 'photoframe', 'coffe_maker', 'food_food', 'ice', 'curtain', 'shaving_cream', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the dishwasher and turn on the faucet. Rinse off the dishes. Put the dishes in the dishwasher. Put detergent in the dishwasher. Close the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <detergent>', '[GRAB] <detergent>', '[PUTBACK] <detergent> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'brush', 'ceilinglamp', 'mat', 'laptop', 'mousepad', 'dining_room', 'wall', 'powersocket', 'food_food', 'oven_mitts', 'kettle', 'mouse', 'cd', 'door', 'hairdryer', 'napkin', 'sink', 'table', 'light']\nTask:Wash monitor\nresponse:\n", "output": {"thoughts": "Clean the grime off of my monitor screen when not in use", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>", "[FIND] <brush>", "[TURNTO] <brush>", "[LOOKAT] <brush>", "[GRAB] <brush>", "[WIPE] <computer>", "[DROP] <brush>", "[FIND] <napkin>", "[GRAB] <napkin>", "[WASH] <napkin>", "[WIPE] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'toilet', 'chair', 'detergent', 'trashcan', 'orchid', 'floor', 'ceiling', 'stamp', 'man', 'bookshelf', 'purse', 'toaster', 'mat', 'oil', 'mousepad', 'nightstand', 'cup', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'curtain', 'mouse', 'headset', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Answer door\nresponse:\n{\n    \"thoughts\": Wait until doorbell rings, look to see who it is outside and open the door if you feel so inclined.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'food_salt', 'measuring_cup', 'toilet', 'jelly', 'drawing', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'alcohol', 'mat', 'shower', 'foundation', 'electric_shaver', 'mousepad', 'wall', 'toy', 'basket_for_clothes', 'photoframe', 'wine', 'dresser', 'laundry_detergent', 'food_steak', 'door', 'napkin', 'bathroom_cabinet', 'dishwasher']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTBACK] <toy> <dresser>", "[CLOSE] <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'food_egg', 'bathroom_counter', 'folder', 'bedroom', 'computer', 'table_cloth', 'piano_bench', 'stovefan', 'tvstand', 'faucet', 'foundation', 'homework', 'couch', 'nightstand', 'powersocket', 'oven', 'basket_for_clothes', 'cup', 'keys', 'food_food', 'mouthwash', 'after_shave', 'doorjamb', 'shoes', 'controller', 'maindoor', 'bag', 'dirt', 'coffee']\nTask:Get ready to leave\nresponse:\n{\n    \"thoughts\": Get your jacket and shoes on, locate the keys off the side table, walk out the front door and secure building with keys when you leave,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>', '[FIND] <keys>', '[GRAB] <keys>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_ice_cream', 'walllamp', 'nail_polish', 'cat', 'orchid', 'drawing', 'floor', 'ceiling', 'purse', 'alcohol', 'mail', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'towel', 'wall', 'oven', 'cup', 'food_cheese', 'television', 'curtain', 'doorjamb', 'door', 'table', 'clothes_shirt']\nTask:Straighten paintings on wall\nresponse:\n", "output": {"thoughts": "Go up to the painting and tilt it which ever way it is uneven until it looks straight.", "actions": ["[WALK] <home_office>", "[WALK] <wall>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'shower', 'window', 'tvstand', 'pillow', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'coffe_maker', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <sink>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'toilet', 'chair', 'bathroom', 'food_carrot', 'cpuscreen', 'shampoo', 'phone', 'ceiling', 'bedroom', 'desk', 'computer', 'toaster', 'shower', 'foundation', 'dustpan', 'photoframe', 'coffe_maker', 'food_rice', 'tray', 'curtain', 'doorjamb', 'console', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher', 'creditcard']\nTask:Put umbrella away\nresponse:\n", "output": {"thoughts": "Place the umbrella in the umbrella bucket point first.", "actions": ["[WALK] <home_office>", "[WALK] <mop_bucket>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'toilet', 'chair', 'freezer', 'detergent', 'food_carrot', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'bedroom', 'piano_bench', 'ceilinglamp', 'mat', 'shower', 'faucet', 'oil', 'food_orange', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'check', 'doorjamb', 'vacuum_cleaner', 'door', 'table', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Put the groceries on the table into the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTBACK] <food_chicken> <freezer>', '[CLOSE] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_carrot> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'crayon', 'cat', 'food_butter', 'trashcan', 'bathroom_counter', 'floor', 'folder', 'mail', 'tvstand', 'window', 'mousepad', 'towel', 'wall', 'nightstand', 'bench', 'toy', 'basket_for_clothes', 'cup', 'food_cheese', 'light_bulb', 'food_food', 'television', 'curtain', 'shoes', 'dresser', 'door', 'sink']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Bend down and pick up the toys off of the floor.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[WALK] <dresser>", "[GRAB] <toy>", "[PUTIN] <toy> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'food_pizza', 'knifeblock', 'food_egg', 'hanger', 'phone', 'desk', 'tea_bag', 'purse', 'envelope', 'clothes_jacket', 'table_cloth', 'faucet', 'window', 'closetdrawer', 'clothes_pants', 'food_jam', 'wall', 'basket_for_clothes', 'bed', 'cd', 'comb', 'laundry_detergent', 'clothes_scarf', 'napkin', 'home_office', 'bathroom_cabinet', 'washing_machine', 'teeth']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[PLUGIN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['beer', 'chair', 'band-aids', 'walllamp', 'bathroom', 'detergent', 'food_carrot', 'drawing', 'floor', 'ceiling', 'mat', 'faucet', 'mousepad', 'food_noodles', 'couch', 'wall', 'basket_for_clothes', 'photoframe', 'light_bulb', 'novel', 'television', 'coffee_filter', 'tray', 'diary', 'mouse', 'pencil', 'cd', 'chessboard', 'bathroom_cabinet', 'creditcard']\nTask:Wipe down baseboards please\nresponse:\n", "output": {"thoughts": "Use a rag to get all the dust off the baseboards please.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <wall>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'food_cheese', 'toothbrush_holder', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'bathroom_cabinet', 'light', 'home_office']\nTask:Work\nresponse:\n{\n    \"thoughts\": Type on a computer in the home office.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'phone', 'character', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'shower', 'faucet', 'window', 'oil', 'couch', 'clothes_pants', 'pillow', 'wall', 'tape', 'cup', 'novel', 'tray', 'curtain', 'mouse', 'chessboard', 'controller', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on a couch, grab a book, open it, and scan the pages", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ground_coffee', 'nail_polish', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'button', 'mail', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'tape', 'television', 'curtain', 'food_onion', 'cupboard', 'table', 'shoe_rack']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to the living room, sit on the couch, find the remote, switch on the TV and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'food_ice_cream', 'tea', 'walllamp', 'drawing', 'bills', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'keyboard', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'iron', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'toy', 'television', 'tray', 'curtain', 'sink', 'bathroom_cabinet', 'table']\nTask:Read magazine\nresponse:\n", "output": {"thoughts": "Please sit on the couch and read a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <newspaper>", "[LOOKAT] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'ground_coffee', 'bathroom', 'trashcan', 'food_butter', 'knifeblock', 'needle', 'floor', 'ceiling', 'character', 'clothes_jacket', 'toaster', 'ceilinglamp', 'balanceball', 'mousepad', 'broom', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'rag', 'coffee_pot', 'laundry_detergent', 'maindoor', 'napkin', 'table', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'jelly', 'chair', 'freezer', 'nail_polish', 'milk', 'hanger', 'floor', 'phone', 'ceiling', 'clothes_jacket', 'piano_bench', 'window', 'oil', 'wall', 'cup', 'keys', 'food_food', 'curtain', 'doorjamb', 'cards', 'pencil', 'vacuum_cleaner', 'dresser', 'pajamas', 'hairdryer', 'door', 'bag', 'bathroom_cabinet']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug in the vacuum. Turn it on. Push it back and forth across the floor until all the debris is gone.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'juice', 'blow_dryer', 'walllamp', 'detergent', 'floor', 'ceiling', 'bedroom', 'keyboard', 'purse', 'tvstand', 'deck_of_cards', 'clothes_pants', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'keys', 'tray', 'kettle', 'shoes', 'hair', 'sink', 'home_office', 'table', 'dishwasher', 'soap', 'board_game', 'dirt']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_bread', 'freezer', 'food_carrot', 'milk', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'mousepad', 'garbage_can', 'clothes_hat', 'food_orange', 'food_dessert', 'wall', 'powersocket', 'food_vegetable', 'food_cheese', 'doorjamb', 'lighter', 'mouse', 'coffee', 'door', 'light', 'colander', 'teeth']\nTask:Rotate stock in refrigerator\nresponse:\n", "output": {"thoughts": "Throw out expired items, put older items in front.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[TURNTO] <food_bread>", "[LOOKAT] <food_bread>", "[PUTOBJBACK] <food_bread>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[FIND] <garbage_can>", "[PUTBACK] <food_cheese> <garbage_can>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[PUTOBJBACK] <food_cheese>", "[FIND] <food_dessert>", "[TURNTO] <food_dessert>", "[LOOKAT] <food_dessert>", "[PUSH] <food_dessert>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <milk>", "[GRAB] <milk>", "[TURNTO] <milk>", "[LOOKAT] <milk>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'microwave', 'freezer', 'ground_coffee', 'dough', 'orchid', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'tea_bag', 'toaster', 'faucet', 'shower', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'curtain', 'kitchen_counter', 'cupboard', 'sink']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['spectacles', 'dog', 'detergent', 'trashcan', 'orchid', 'needle', 'bathroom_counter', 'computer', 'button', 'table_cloth', 'deck_of_cards', 'stereo', 'shower', 'floor_lamp', 'towel', 'wall', 'love_seat', 'cup', 'newspaper', 'basket_for_clothes', 'photoframe', 'food_rice', 'wooden_spoon', 'doorjamb', 'dresser', 'controller', 'door', 'napkin', 'home_office', 'clothes_skirt']\nTask:Read\nresponse:\n", "output": {"thoughts": "Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'food_pizza', 'tablelamp', 'microwave', 'floor', 'ceiling', 'desk', 'bookshelf', 'stovefan', 'mat', 'ceilinglamp', 'laptop', 'window', 'homework', 'oil', 'closetdrawer', 'couch', 'nightstand', 'wall', 'cup', 'bed', 'food_food', 'mouthwash', 'toothbrush_holder', 'oven_mitts', 'razor', 'food_donut', 'check', 'clothes_scarf', 'bathroom_cabinet']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Walk to bedroom. Walk to bed, lie in bed.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['glue', 'crayon', 'freezer', 'nail_polish', 'food_carrot', 'milk', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'cup', 'bed', 'food_food', 'doorjamb', 'mouse', 'pajamas', 'controller', 'door', 'spoon', 'napkin', 'light', 'dirt']\nTask:Get glass of milk\nresponse:\n", "output": {"thoughts": "Get an empty glass. Take milk from refrigerator and open it. Pour milk into glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'cards', 'dresser', 'door', 'colander']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'tablelamp', 'toilet', 'jelly', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'coffee_filter', 'curtain', 'doorjamb', 'centerpiece', 'door', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit down on sofa, get the remote, turn on tv, change channel to show I want to watch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'freezer', 'orchid', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'phone', 'bookshelf', 'stovefan', 'pillow', 'wall', 'wall_clock', 'cutting_board', 'box', 'shelf', 'newspaper', 'bed', 'novel', 'coffe_maker', 'toilet_paper', 'wooden_spoon', 'hairbrush', 'door', 'home_office', 'table', 'light', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Pick up the book, open to current page, read the words.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['ironing_board', 'dry_pasta', 'food_pizza', 'beer', 'dog', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'phone', 'folder', 'character', 'keyboard', 'stovefan', 'laptop', 'iron', 'picture', 'mousepad', 'microphone', 'powersocket', 'cup', 'novel', 'food_food', 'food_donut', 'wine', 'pencil', 'bag', 'table', 'soap', 'clothes_shirt']\nTask:Iron shirt\nresponse:\n", "output": {"thoughts": "Set up an ironing board and an iron.  Lay a shirt on board.  Run iron over shirt until it is free of wrinkles.", "actions": ["[WALK] <bedroom>", "[WALK] <ironing_board>", "[FIND] <ironing_board>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <ironing_board>", "[FIND] <iron>", "[GRAB] <iron>", "[PLUGIN] <iron>", "[PLUGOUT] <iron>", "[PUTOBJBACK] <iron>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'rag', 'doorjamb', 'cards', 'dresser', 'sink', 'dishwasher']\nTask:Watch  TV\nresponse:\n{\n    \"thoughts\": Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_ice_cream', 'tea', 'knifeblock', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'envelope', 'alcohol', 'mat', 'faucet', 'shower', 'towel', 'wall', 'toy', 'bed', 'wooden_spoon', 'doorjamb', 'vacuum_cleaner', 'pajamas', 'laundry_detergent', 'door', 'sink', 'bathroom_cabinet', 'light', 'clothes_skirt', 'soap', 'hands_both']\nTask:Hand washing\nresponse:\n", "output": {"thoughts": "Turn on the faucet. Soap up your hands. Then rinse under the running water of the faucet. Turn off faucet and dry hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'food_ice_cream', 'toilet', 'walllamp', 'detergent', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'tvstand', 'laptop', 'ceilinglamp', 'brush', 'couch', 'pillow', 'dining_room', 'wall', 'child', 'oven', 'photoframe', 'bed', 'television', 'doorjamb', 'mouse', 'dishwasher']\nTask:Let baby learn how to walk\nresponse:\n{\n    \"thoughts\": The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[GREET] <child>', '[FIND] <couch>', '[TURNTO] <couch>', '[POINTAT] <couch>', '[TURNTO] <child>', '[LOOKAT] <child>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'scissors', 'microwave', 'chair', 'freezer', 'food_butter', 'drawing', 'knife', 'cpuscreen', 'floor', 'folder', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'shower', 'mousepad', 'wall', 'nightstand', 'bed', 'ice', 'after_shave', 'doorjamb', 'mouse', 'kitchen_counter', 'food_steak', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I work on my HP computer.", "actions": ["[WALK] <bedroom>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[WALK] <bed>", "[SIT] <bed>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'mop', 'microwave', 'toilet', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'computer', 'piano_bench', 'ceilinglamp', 'plate', 'deck_of_cards', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'bed', 'novel', 'television', 'doorjamb', 'door', 'coffee_cup', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'walllamp', 'nail_polish', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bookshelf', 'computer', 'faucet', 'oil', 'dining_room', 'wall', 'nightstand', 'oven', 'powersocket', 'toy', 'cup', 'coffee_filter', 'shaving_cream', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'home_office', 'table', 'dishwasher', 'bathroom_cabinet']\nTask:Get glass of water\nresponse:\n", "output": {"thoughts": "Get a glass from a cabinet. Hold it under the sink and turn the cold water on. Fill the glass. Turn off the water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <kitchen_cabinet>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'tablelamp', 'chair', 'dog', 'walllamp', 'ceilingfan', 'knifeblock', 'needle', 'bathroom_counter', 'floor', 'computer', 'shoe-shine_kit', 'dustpan', 'box', 'cup', 'bed', 'keys', 'television', 'food_donut', 'tray', 'curtain', 'doorjamb', 'kettle', 'kitchen_counter', 'shoes', 'comb', 'sink', 'bathroom_cabinet', 'board_game']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": go to the bathroom and brush your teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'toothbrush', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'shampoo', 'floor', 'ceiling', 'desk', 'mop_bucket', 'stovefan', 'laptop', 'tvstand', 'faucet', 'wall', 'nightstand', 'oven', 'tape', 'food_food', 'tray', 'curtain', 'kettle', 'kitchen_counter', 'cupboard', 'conditioner', 'sink', 'bathroom_cabinet', 'table']\nTask:Wipe down counter\nresponse:\n", "output": {"thoughts": "Clean counter", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <rag>", "[SQUEEZE] <rag>", "[SWITCHOFF] <faucet>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'comb', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office, walk to desk, sit in chair, turn on computer, look at computer, click mouse and type on keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'food_bread', 'dog', 'ground_coffee', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'bedroom', 'alcohol', 'mop_bucket', 'stovefan', 'laptop', 'tvstand', 'ceilinglamp', 'shoe-shine_kit', 'window', 'mat', 'mousepad', 'food_orange', 'pillow', 'powersocket', 'coffe_maker', 'fly', 'coffee_pot', 'kitchen_counter', 'cupboard', 'dishwasher', 'soap']\nTask:Watch fly\nresponse:\n", "output": {"thoughts": "Look around for a fly, often on a wall or ceiling.  Follow it's movements.", "actions": ["[WALK] <bedroom>", "[WALK] <fly>", "[FIND] <fly>", "[TURNTO] <fly>", "[LOOKAT] <fly>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'ground_coffee', 'walllamp', 'knife', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bedroom', 'button', 'stovefan', 'mat', 'shower', 'clothes_socks', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'bed', 'light_bulb', 'check', 'video_game_controller', 'mouthwash', 'food_onion', 'vacuum_cleaner', 'pajamas', 'cupboard', 'hair', 'instrument_guitar']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go to bedroom, put pajamas on, lie in bed, sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <pajamas>', '[FIND] <pajamas>', '[GRAB] <pajamas>', '[PUTON] <pajamas>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'detergent', 'food_carrot', 'orchid', 'drawing', 'needle', 'hanger', 'floor', 'computer', 'bookmark', 'mat', 'laptop', 'faucet', 'shower', 'couch', 'candle', 'wall', 'nightstand', 'pasta', 'bed', 'photoframe', 'toilet_paper', 'toothbrush_holder', 'tray', 'kitchen_counter', 'controller', 'spoon', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I lie on my bed and read a book.", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'bathroom', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'food_egg', 'bathroom_counter', 'shampoo', 'phone', 'mat', 'ceilinglamp', 'shower', 'electric_shaver', 'clothes_socks', 'couch', 'microphone', 'dining_room', 'food_kiwi', 'nightstand', 'wall', 'check', 'food_food', 'after_shave', 'mouse', 'kitchen_counter', 'cupboard', 'home_office', 'laser_pointer']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'band-aids', 'walllamp', 'drawing', 'food_egg', 'bathroom_counter', 'ceiling', 'stamp', 'desk', 'cleaning_solution', 'computer', 'stovefan', 'deck_of_cards', 'faucet', 'tvstand', 'window', 'couch', 'wall', 'nightstand', 'coffe_maker', 'television', 'mirror', 'after_shave', 'paper_towel', 'clothes_dress', 'sink', 'bag', 'bathroom_cabinet', 'dishwasher', 'laser_pointer']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[CLOSE] <bathroom_cabinet>", "[WALK] <mirror>", "[FIND] <cleaning_solution>", "[RINSE] <cleaning_solution>", "[WIPE] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'ground_coffee', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'bedroom', 'ceilinglamp', 'tvstand', 'stereo', 'food_orange', 'cup', 'bed', 'keys', 'novel', 'food_food', 'coffe_maker', 'mouthwash', 'coin', 'check', 'diary', 'coffee_pot', 'mouse', 'dresser', 'hair', 'table', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to my bedroom, and then I switch on the radio,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'video_game_console', 'desk', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'video_game_controller', 'television', 'food_rice', 'ice', 'curtain', 'doorjamb', 'dresser', 'door', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on game.  Use controller", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[PLUGIN] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'crayon', 'microwave', 'toilet', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'shampoo', 'computer', 'mat', 'tvstand', 'picture', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'bathroom_cabinet', 'light', 'instrument_violin']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk over to the phone. .I reach my arm out. I open my hand. I place my hand on the phone. I close my fingers around the phone. I pick up the phone.,\n    \"actions\": ['[WALK] <phone>', '[TOUCH] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'filing_cabinet', 'microwave', 'toilet', 'dog', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'folder', 'keyboard', 'thread', 'computer', 'mat', 'window', 'dining_room', 'wall', 'nightstand', 'bench', 'cup', 'wine', 'curtain', 'kitchen_counter', 'hairbrush', 'clothes_dress', 'food_steak']\nTask:Look out window\nresponse:\n", "output": {"thoughts": "I head into my child's room and look out the window to view the neighborhood from a different area in the house.", "actions": ["[WALK] <bedroom>", "[WALK] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'toilet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'mail', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cutting_board', 'photoframe', 'bed', 'television', 'doorjamb', 'dishrack', 'dresser', 'door', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": go to the living room to watch TV,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[TOUCH] <remote_control>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[DROP] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'sheets', 'scissors', 'microwave', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'toothbrush', 'floor', 'phone', 'toaster', 'mat', 'tvstand', 'dustpan', 'wall', 'box', 'photoframe', 'coffe_maker', 'check', 'oven_mitts', 'food_rice', 'curtain', 'coffee_pot', 'kitchen_counter', 'cupboard', 'hairdryer', 'dishwasher', 'soap', 'coffee']\nTask:Take dishes out of dishwasher\nresponse:\n", "output": {"thoughts": "Take clean dishes out of the dishwasher and put away", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I walk into my home office and walk up to my desk. I pull out the chair and sit down. I turn on the computer. After the computer is turned on I click on the google icon. I enter Disney Cruise Line into the search bar and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[TURNTO] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TURNTO] <keyboard>', '[LOOKAT] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'dog', 'trashcan', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'stamp', 'bookshelf', 'thread', 'computer', 'food_bacon', 'brush', 'ceilinglamp', 'shower', 'window', 'mousepad', 'food_jam', 'pillow', 'wall', 'check', 'curtain', 'doorjamb', 'shoes', 'dresser', 'light', 'laser_pointer']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Using soft brush, carefully dust both Indonesian wedding doll statues.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <brush>", "[GRAB] <brush>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'blender', 'freezer', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'orchid', 'hanger', 'floor', 'bedroom', 'keyboard', 'computer', 'stereo', 'window', 'oil', 'couch', 'pillow', 'wall', 'cup', 'bed', 'food_donut', 'curtain', 'centerpiece', 'console', 'shoes', 'dresser', 'maindoor', 'home_office', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Turn on radio, put on favorite channel, listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'freezer', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'alcohol', 'stovefan', 'brush', 'tvstand', 'picture', 'dustpan', 'wall', 'powersocket', 'cup', 'keys', 'light_bulb', 'food_food', 'doorjamb', 'controller', 'door', 'clothes_scarf', 'bathroom_cabinet', 'light', 'dishwasher', 'home_office', 'clothes_shirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Hit light switch", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'cd_player', 'knife', 'character', 'tea_bag', 'remote_control', 'button', 'table_cloth', 'ceilinglamp', 'shoe-shine_kit', 'shower', 'food_peanut_butter', 'candle', 'nightstand', 'bench', 'cutting_board', 'shelf', 'bed', 'photoframe', 'oven_mitts', 'food_rice', 'doorjamb', 'shaving_cream', 'mouse', 'dresser', 'controller', 'sink', 'table', 'clothes_skirt']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[PLUGIN] <cd_player>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['food_salt', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'bookshelf', 'mail', 'table_cloth', 'shower', 'foundation', 'dustpan', 'microphone', 'wall', 'powersocket', 'food_food', 'check', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'sink', 'hairdryer', 'light', 'clothes_skirt', 'scrabble', 'creditcard']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Hit the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'blender', 'walllamp', 'ceilingfan', 'dough', 'drawing', 'floor', 'ceiling', 'desk', 'stamp', 'remote_control', 'food_cake', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'wall', 'oven', 'photoframe', 'novel', 'television', 'curtain', 'home_office', 'table', 'instrument_guitar']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch nd watch t.V,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'food_bread', 'microwave', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'bathroom_counter', 'bedroom', 'tea_bag', 'tvstand', 'window', 'form', 'pillow', 'broom', 'bench', 'basket_for_clothes', 'newspaper', 'bed', 'check', 'coffe_maker', 'tray', 'door', 'home_office', 'light', 'bathroom_cabinet', 'tooth_paste', 'coffee']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "I get dressed in my bedroom in the mornings.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'freezer', 'detergent', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'bookshelf', 'mail', 'toaster', 'ceilinglamp', 'faucet', 'window', 'pillow', 'dining_room', 'wall', 'powersocket', 'oven', 'novel', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'pencil', 'vacuum_cleaner', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on light in an area,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'food_ice_cream', 'chair', 'band-aids', 'toothbrush', 'orchid', 'needle', 'hanger', 'bedroom', 'desk', 'bookshelf', 'piano_bench', 'ceilinglamp', 'mat', 'clothes_socks', 'microphone', 'dining_room', 'bowl', 'nightstand', 'tray', 'curtain', 'diary', 'dresser', 'cupboard', 'spoon', 'bathroom_cabinet', 'table', 'dishwasher', 'light']\nTask:Open door\nresponse:\n", "output": {"thoughts": "I walk up to the door.  I use my hand to turn the door knob to the left.  I pull the door open with the same hand I used to turn the door knob", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'wallshelf', 'clothes_underwear', 'floor', 'desk', 'mail', 'button', 'stovefan', 'toaster', 'mousepad', 'soap', 'pillow', 'dining_room', 'photoframe', 'television', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'controller', 'bathtub', 'bathroom_cabinet', 'colander', 'clothes_shirt', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find a cat in the kitchen and touch it,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'band-aids', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'envelope', 'computer', 'ceilinglamp', 'tvstand', 'stereo', 'form', 'wall', 'nightstand', 'powersocket', 'cup', 'bed', 'photoframe', 'coffe_maker', 'check', 'centerpiece', 'dresser', 'food_steak', 'napkin', 'tooth_paste', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Put iPhone into speaker holder, open up the Pandora app, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <stereo>", "[PLUGIN] <phone>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'food_carrot', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'mail', 'ceilinglamp', 'mat', 'iron', 'balanceball', 'faucet', 'food_orange', 'pillow', 'wall', 'pasta', 'box', 'cup', 'newspaper', 'food_food', 'check', 'toilet_paper', 'chessboard', 'maindoor', 'napkin', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Putting groceries in refridgerator,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'orchid', 'bills', 'cpuscreen', 'knifeblock', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'microphone', 'form', 'towel', 'wall', 'oven', 'food_food', 'television', 'ice', 'cards', 'mouse', 'pencil', 'cupboard', 'door', 'clothes_scarf', 'table']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Wipe down computer desk and monitor", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <desk>", "[FIND] <computer>", "[WIPE] <computer>", "[WIPE] <desk>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['band-aids', 'trashcan', 'wallshelf', 'clothes_underwear', 'desk', 'keyboard', 'mail', 'computer', 'tvstand', 'shower', 'window', 'clothes_socks', 'couch', 'food_peanut_butter', 'pillow', 'basket_for_clothes', 'keys', 'coffe_maker', 'television', 'food_donut', 'tray', 'rag', 'headset', 'clothes_scarf', 'bathtub', 'sink', 'light', 'dishwasher', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'mop', 'bathroom', 'knife', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'desk', 'food_cake', 'window', 'electric_shaver', 'microphone', 'food_jam', 'dining_room', 'wall', 'photoframe', 'food_food', 'television', 'curtain', 'mouse', 'kitchen_counter', 'cd', 'chessboard', 'dresser', 'hairbrush', 'clothes_scarf', 'bathroom_cabinet', 'dishwasher', 'instrument_violin']\nTask:Put on coat\nresponse:\n", "output": {"thoughts": "I remove my scarf from a coat peg in the entrance hall.  I wrap the scarf around my neck", "actions": ["[WALK] <home_office>", "[WALK] <hanger>", "[TURNTO] <hanger>", "[LOOKAT] <hanger>", "[FIND] <clothes_scarf>", "[GRAB] <clothes_scarf>", "[PUTON] <clothes_scarf>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'microwave', 'freezer', 'walllamp', 'bathroom', 'toothbrush', 'orchid', 'hanger', 'longboard', 'phone', 'desk', 'computer', 'food_bacon', 'table_cloth', 'clothes_jacket', 'brush', 'ceilinglamp', 'faucet', 'foundation', 'mousepad', 'clothes_pants', 'pasta', 'light_bulb', 'toothbrush_holder', 'coin', 'coffee_filter', 'towel_rack', 'hair', 'vase', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'jelly', 'chair', 'trashcan', 'woman', 'orchid', 'bathroom_counter', 'thread', 'mail', 'toaster', 'mousepad', 'clothes_socks', 'candle', 'towel', 'dining_room', 'broom', 'pasta', 'powersocket', 'oven', 'food_food', 'after_shave', 'pencil', 'hair', 'hairdryer', 'door', 'napkin', 'home_office', 'colander', 'shoe_rack']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet people into my home", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I need to write an e-mail to my boss about the meeting at work tomorrow.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[TYPE] <keyboard>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'bathroom', 'detergent', 'food_carrot', 'orchid', 'drawing', 'cpuscreen', 'food_egg', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'brush', 'ceilinglamp', 'tvstand', 'mat', 'stereo', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'basket_for_clothes', 'television', 'curtain', 'doorjamb', 'door', 'sink', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on my couch and watch TV in my living room", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'cat', 'ceilingfan', 'food_butter', 'drawing', 'longboard', 'floor', 'ceiling', 'shampoo', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'couch', 'clothes_pants', 'wall', 'wall_clock', 'bench', 'television', 'curtain', 'kettle', 'towel_rack', 'hair', 'pajamas', 'maindoor', 'home_office', 'table', 'vase']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Changing channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[DROP] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'ground_coffee', 'detergent', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'bookshelf', 'envelope', 'mat', 'shower', 'faucet', 'homework', 'mousepad', 'wall', 'toy', 'photoframe', 'check', 'coffee_filter', 'food_onion', 'shaving_cream', 'cards', 'mouse', 'dresser', 'hairbrush', 'door', 'soap', 'scrabble']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Enter inside kid's room, begin picking up toys from the floor and putting them in a toy box, leave once once.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'dough', 'drawing', 'orchid', 'cpuscreen', 'food_egg', 'hanger', 'floor', 'keyboard', 'bookshelf', 'food_bacon', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'cup', 'novel', 'check', 'curtain', 'doorjamb', 'diary', 'pencil', 'kitchen_counter', 'bag', 'home_office', 'table', 'dishwasher', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'food_carrot', 'knife', 'drawing', 'bills', 'floor', 'desk', 'computer', 'food_bacon', 'bookmark', 'homework', 'candle', 'dining_room', 'wall', 'nightstand', 'toy', 'basket_for_clothes', 'bed', 'check', 'coffe_maker', 'television', 'mouse', 'hair', 'controller', 'conditioner', 'sink', 'spoon', 'home_office']\nTask:Breakfast\nresponse:\n", "output": {"thoughts": "I have breakfast in bed, watching tv.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[GRAB] <food_food>", "[WALK] <bedroom>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <bed>", "[LIE] <bed>", "[TURNTO] <television>", "[WATCH] <television>", "[EAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'water_glass', 'pot', 'glue', 'walllamp', 'trashcan', 'knifeblock', 'ceiling', 'clothes_jacket', 'stovefan', 'toaster', 'mat', 'faucet', 'microphone', 'wall', 'cup', 'bed', 'food_food', 'check', 'water', 'food_donut', 'curtain', 'kitchen_counter', 'chessboard', 'cupboard', 'laundry_detergent', 'sink', 'spoon', 'bathroom_cabinet', 'dirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['walllamp', 'nail_polish', 'bathroom', 'detergent', 'knife', 'cpuscreen', 'dvd_player', 'bedroom', 'computer', 'table_cloth', 'mop_bucket', 'brush', 'toaster', 'shower', 'window', 'couch', 'dining_room', 'wall', 'pasta', 'cup', 'photoframe', 'after_shave', 'pencil', 'kitchen_counter', 'dresser', 'clothes_dress', 'sink', 'instrument_guitar', 'light', 'creditcard']\nTask:Sit\nresponse:\n", "output": {"thoughts": "We sit at the dining room table", "actions": ["[WALK] <dining_room>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_salt', 'chair', 'jelly', 'walllamp', 'milk', 'floor', 'bedroom', 'keyboard', 'button', 'plate', 'tvstand', 'shower', 'window', 'dustpan', 'dining_room', 'nightstand', 'oven', 'basket_for_clothes', 'check', 'coffe_maker', 'tray', 'controller', 'clothes_dress', 'food_steak', 'napkin', 'soap', 'clothes_shirt', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'crayon', 'toilet', 'ground_coffee', 'drawing', 'bills', 'cpuscreen', 'clothes_underwear', 'orchid', 'floor', 'phone', 'ceiling', 'bookshelf', 'mat', 'shower', 'foundation', 'homework', 'dining_room', 'wall', 'tape', 'cutting_board', 'photoframe', 'newspaper', 'food_food', 'toothbrush_holder', 'curtain', 'kitchen_counter', 'soap']\nTask:Put clothes away\nresponse:\n", "output": {"thoughts": "I take clean clothes and either place them in dresser drawers or hang them in a closet.", "actions": ["[WALK] <bedroom>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'cd_player', 'walllamp', 'trashcan', 'food_carrot', 'drawing', 'needle', 'bathroom_counter', 'dvd_player', 'bedroom', 'bookshelf', 'alcohol', 'piano_bench', 'bookmark', 'laptop', 'stereo', 'couch', 'pillow', 'broom', 'bench', 'basket_for_clothes', 'mouse', 'shoes', 'headset', 'food_steak', 'sink', 'hairdryer', 'instrument_guitar', 'door', 'bag']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the playback device you want to listen to. turn on. press play. enjoyu music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['food_bread', 'food_salt', 'beer', 'walllamp', 'drawing', 'orchid', 'knifeblock', 'clothes_underwear', 'floor', 'folder', 'man', 'computer', 'toaster', 'ceilinglamp', 'stereo', 'window', 'homework', 'mousepad', 'couch', 'microphone', 'pillow', 'wall', 'nightstand', 'keys', 'food_food', 'wooden_spoon', 'curtain', 'doorjamb', 'hair', 'table']\nTask:Have conversation with boyfriend\nresponse:\n", "output": {"thoughts": "Sit on sofa, talk", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'detergent', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'purse', 'toaster', 'mat', 'faucet', 'iron', 'mousepad', 'wall', 'powersocket', 'cup', 'coffe_maker', 'water', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'clothes_scarf', 'colander']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'shower', 'mousepad', 'pillow', 'dining_room', 'wall', 'wall_clock', 'nightstand', 'tape', 'bed', 'console', 'mouse', 'clothes_scarf', 'light', 'soap', 'teeth']\nTask:Check email\nresponse:\n", "output": {"thoughts": "Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[WALK] <chair>", "[SIT] <chair>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'toilet', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'plate', 'faucet', 'drying_rack', 'dish_soap', 'couch', 'dining_room', 'wall', 'oven', 'newspaper', 'photoframe', 'check', 'toilet_paper', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'table']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": I walk in to the Kitchen. I find a plate, a sponge, and the dish soap. I find the faucet. I turn the faucet on. I pour water on the sponge and plate. I add dish soap to the sponge. I put the dish soap back. I scrub the plate. I rinse the plate. I place the plate on the drying rack. I rinse the sponge. I turn the faucet off.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <sponge>', '[GRAB] <sponge>', '[RINSE] <sponge>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[WASH] <plate>', '[RINSE] <plate>', '[FIND] <drying_rack>', '[PUTBACK] <plate> <plate>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'walllamp', 'woman', 'knifeblock', 'hanger', 'ceiling', 'folder', 'character', 'man', 'mop_bucket', 'bookmark', 'shower', 'window', 'electric_shaver', 'pillow', 'dining_room', 'wall_clock', 'cutting_board', 'bed', 'television', 'wine', 'curtain', 'mouse', 'pencil', 'sink', 'home_office', 'light', 'dirt', 'shoe_rack']\nTask:Greet people\nresponse:\n", "output": {"thoughts": "Say HI to those that come in.", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'beer', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'stovefan', 'mat', 'toaster', 'shower', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'food_food', 'food_donut', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Walk into the kitchen. Open the fridge. Take out the chicken. Open the kitchen cabinet. Take out the frying pan. Put the chicken in the frying pan. Put the frying pan on the stove. Turn on the stove. Cook the chicken.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[CLOSE] <freezer>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[CLOSE] <kitchen_cabinet>', '[WALK] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'basket_for_clothes', 'food_cheese', 'television', 'curtain', 'doorjamb', 'cd', 'conditioner', 'door', 'clothes_scarf', 'home_office', 'table', 'dishwasher', 'instrument_guitar']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch nd watch t.V", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'trashcan', 'orchid', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'window', 'food_noodles', 'couch', 'pillow', 'wall', 'photoframe', 'toothbrush_holder', 'curtain', 'doorjamb', 'chessboard', 'door', 'bathtub', 'bag', 'home_office', 'light', 'table', 'dishwasher', 'instrument_violin']\nTask:Admire art\nresponse:\n{\n    \"thoughts\": Looking at painting, art,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'toilet', 'chair', 'freezer', 'band-aids', 'bathroom', 'food_carrot', 'drawing', 'floor', 'mat', 'window', 'couch', 'clothes_hat', 'form', 'pillow', 'bench', 'food_food', 'check', 'coffe_maker', 'toilet_paper', 'ice', 'centerpiece', 'doorjamb', 'cupboard', 'home_office', 'table', 'light', 'bathroom_cabinet', 'scrabble']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_salt', 'toilet', 'chair', 'bathroom', 'trashcan', 'drawing', 'floor', 'character', 'desk', 'purse', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'foundation', 'couch', 'pillow', 'nightstand', 'powersocket', 'novel', 'tray', 'mouse', 'bathroom_cabinet', 'soap', 'teeth']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cat', 'food_carrot', 'toothbrush', 'drawing', 'orchid', 'floor', 'mail', 'stovefan', 'ceilinglamp', 'window', 'mousepad', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'photoframe', 'mouthwash', 'tray', 'curtain', 'coffee_pot', 'pajamas', 'door', 'sink', 'bathroom_cabinet', 'table', 'light', 'tooth_paste']\nTask:Relax\nresponse:\n", "output": {"thoughts": "Lie down on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'beer', 'chair', 'freezer', 'knifeblock', 'bathroom_counter', 'longboard', 'phone', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'balanceball', 'stereo', 'mousepad', 'clothes_pants', 'nightstand', 'bench', 'bed', 'photoframe', 'novel', 'television', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'food_steak', 'sink', 'hairdryer', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the book shelf, get a book, open the book and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tea', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'pasta', 'powersocket', 'coffe_maker', 'doorjamb', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'home_office', 'table', 'tooth_paste']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Push power button.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'woman', 'knifeblock', 'needle', 'cpuscreen', 'desk', 'man', 'keyboard', 'bedroom', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'nightstand', 'oven', 'cup', 'food_food', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['blender', 'detergent', 'orchid', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'purse', 'envelope', 'mat', 'window', 'foundation', 'broom', 'wall', 'nightstand', 'keys', 'novel', 'check', 'toothbrush_holder', 'oven_mitts', 'curtain', 'doorjamb', 'mouse', 'dresser', 'hairdryer', 'home_office', 'clothes_shirt', 'coffee', 'creditcard']\nTask:Put on coat\nresponse:\n", "output": {"thoughts": "Open closet and grab coat.  Put on coat and fasten the front closed.  Close closet.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'blender', 'food_ice_cream', 'food_butter', 'phone', 'character', 'bookshelf', 'brush', 'ceilinglamp', 'plate', 'shower', 'iron', 'stereo', 'toaster', 'oil', 'food_orange', 'pillow', 'cup', 'food_food', 'check', 'oven_mitts', 'food_donut', 'food_onion', 'after_shave', 'mouse', 'hairdryer', 'bag', 'bathroom_cabinet', 'home_office']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <home_office>', '[DROP] <plate>', '[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <home_office>', '[DROP] <cup>', '[WALK] <dining_room>', '[WALK] <fork>', '[FIND] <fork>', '[GRAB] <fork>', '[WALK] <home_office>', '[DROP] <fork>']\n}\n", "input": "\nScene_Objects:['chair', 'band-aids', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'mousepad', 'oil', 'broom', 'wall', 'nightstand', 'bench', 'newspaper', 'coffe_maker', 'food_food', 'coin', 'food_donut', 'lighter', 'mouse', 'chessboard', 'door', 'table', 'dirt']\nTask:Turking\nresponse:\n", "output": {"thoughts": "Do a random variety very small jobs from different clients on mturk.Com", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'dog', 'cat', 'drawing', 'wallshelf', 'ceiling', 'folder', 'character', 'cleaning_solution', 'envelope', 'computer', 'brush', 'toaster', 'iron', 'mousepad', 'closetdrawer', 'couch', 'bed', 'photoframe', 'novel', 'coffe_maker', 'centerpiece', 'hair', 'cupboard', 'sink', 'home_office', 'light', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'food_salt', 'cd_player', 'chair', 'freezer', 'bathroom', 'food_carrot', 'orchid', 'knifeblock', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'remote_control', 'computer', 'window', 'mousepad', 'wall', 'nightstand', 'light_bulb', 'check', 'food_onion', 'curtain', 'mouse', 'door', 'light', 'tooth_paste']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.", "actions": ["[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'toilet', 'bathroom', 'detergent', 'ceilingfan', 'food_butter', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'pillow', 'wall', 'food_food', 'television', 'tray', 'curtain', 'after_shave', 'cupboard', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": TURN ON TV, SIT ON COUCH ACROSS THE ROOM, USE REMOTE TO LOOK FOR SHOW TO WATCH,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'toilet', 'jelly', 'walllamp', 'drawing', 'orchid', 'needle', 'bathroom_counter', 'floor', 'folder', 'keyboard', 'toaster', 'ceilinglamp', 'window', 'food_noodles', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'light_bulb', 'television', 'curtain', 'food_onion', 'chessboard', 'cupboard', 'door', 'napkin', 'instrument_guitar', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_socks', 'closetdrawer', 'mousepad', 'clothes_pants', 'wall', 'powersocket', 'oven', 'cup', 'newspaper', 'food_food', 'check', 'coffe_maker', 'shaving_cream', 'mouse', 'pencil', 'shoes', 'dresser', 'hair', 'conditioner', 'light', 'clothes_shirt', 'shoe_rack']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Open drawer. Pull out pants and put them on. Take a shirt out and put it on. Put on socks and shoes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTON] <clothes_socks>', '[CLOSE] <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'food_pizza', 'tablelamp', 'microwave', 'toilet', 'freezer', 'bathroom', 'drawing', 'needle', 'floor', 'shampoo', 'desk', 'bookshelf', 'shower', 'food_noodles', 'wall', 'powersocket', 'cup', 'photoframe', 'novel', 'food_donut', 'tray', 'curtain', 'doorjamb', 'shaving_cream', 'pencil', 'door', 'home_office']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I go to the bathroom. I locate the toilet. I walk to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[WALK] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_carrot', 'drawing', 'floor', 'character', 'computer', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'candle', 'wall', 'wall_clock', 'oven', 'cup', 'photoframe', 'bed', 'novel', 'check', 'toilet_paper', 'television', 'curtain', 'towel_rack', 'mouse', 'hairbrush', 'laundry_detergent', 'table', 'clothes_skirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the couch in the living room, sit on ti and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'dough', 'drawing', 'orchid', 'cpuscreen', 'food_egg', 'hanger', 'floor', 'keyboard', 'bookshelf', 'food_bacon', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'cup', 'novel', 'check', 'curtain', 'doorjamb', 'diary', 'pencil', 'kitchen_counter', 'bag', 'home_office', 'table', 'dishwasher', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'trashcan', 'ceilingfan', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'laptop', 'balanceball', 'mousepad', 'pillow', 'wall', 'cup', 'bed', 'television', 'razor', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'spoon', 'light', 'vase']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": push any ligth button to turn it on, for example in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'walllamp', 'orchid', 'drawing', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'iron', 'window', 'shower', 'couch', 'pillow', 'food_kiwi', 'wall', 'pasta', 'powersocket', 'bed', 'check', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'controller', 'door', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to Living room. Find Television. Walk to sofa. Sit on sofa. Find Remote. Grab remote. Turn to television. Look at television. Switch on Television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_salt', 'jelly', 'walllamp', 'detergent', 'knife', 'drawing', 'hanger', 'floor', 'dvd_player', 'character', 'desk', 'bedroom', 'plate', 'dish_soap', 'oil', 'wall', 'powersocket', 'bed', 'light_bulb', 'coffe_maker', 'toilet_paper', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put dishes in the dishwasher and start the dishwasher.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[WALK] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'bathroom', 'milk', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'envelope', 'computer', 'mat', 'ceilinglamp', 'stereo', 'mousepad', 'wall', 'toothbrush_holder', 'food_rice', 'ice', 'after_shave', 'mouse', 'headset', 'dresser', 'door', 'home_office', 'table', 'board_game']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office, walk to desk, sit in chair, turn on computer, look at computer, click mouse and type on keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'tea', 'microwave', 'walllamp', 'detergent', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'stovefan', 'toaster', 'faucet', 'stereo', 'bowl', 'food_kiwi', 'wall', 'oven', 'bed', 'coffe_maker', 'food_food', 'drinking_glass', 'food_donut', 'tray', 'after_shave', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'clothes_skirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'juice', 'microwave', 'toilet', 'chair', 'freezer', 'ground_coffee', 'bathroom', 'knife', 'ceiling', 'bedroom', 'clothes_jacket', 'bookmark', 'toaster', 'faucet', 'iron', 'couch', 'food_peanut_butter', 'dining_room', 'oven', 'cup', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'home_office']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find a cat, touch it gently", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'scissors', 'microwave', 'chair', 'freezer', 'food_butter', 'drawing', 'knife', 'cpuscreen', 'floor', 'folder', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'shower', 'mousepad', 'wall', 'nightstand', 'bed', 'ice', 'after_shave', 'doorjamb', 'mouse', 'kitchen_counter', 'food_steak', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I work on my HP computer.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[WALK] <bed>', '[SIT] <bed>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'walllamp', 'food_butter', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'stereo', 'faucet', 'window', 'mousepad', 'pillow', 'wall', 'powersocket', 'bed', 'coffee_table', 'television', 'doorjamb', 'pencil', 'dresser', 'clothes_dress', 'door', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room and look at the coffee table. Grab the remote control and turn to the television and point the remote control. Put back the remote control.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <coffee_table>", "[TURNTO] <coffee_table>", "[LOOKAT] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'measuring_cup', 'chair', 'nail_polish', 'cat', 'drawing', 'hanger', 'floor', 'phone', 'desk', 'keyboard', 'food_bacon', 'table_cloth', 'bookmark', 'stovefan', 'plate', 'shower', 'dish_soap', 'couch', 'wall', 'powersocket', 'basket_for_clothes', 'toilet_paper', 'pencil', 'hair', 'home_office', 'light', 'dishwasher', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'nail_polish', 'bathroom', 'cat', 'detergent', 'trashcan', 'milk', 'orchid', 'knife', 'computer', 'food_bacon', 'faucet', 'stereo', 'oil', 'food_noodles', 'microphone', 'dining_room', 'nightstand', 'oven', 'food_cheese', 'novel', 'check', 'coffe_maker', 'food_onion', 'coffee_pot', 'kitchen_counter', 'dresser', 'hairdryer', 'home_office']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to bedroom, sit down, pick up book, read", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'measuring_cup', 'microwave', 'chair', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'cleaning_solution', 'keyboard', 'mop_bucket', 'stovefan', 'faucet', 'couch', 'clothes_pants', 'coffe_maker', 'mirror', 'curtain', 'rag', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'dishwasher']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'toilet', 'bathroom', 'milk', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'towel', 'wall', 'bench', 'cup', 'photoframe', 'toothbrush_holder', 'water', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'clothes_skirt', 'soap']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[TURNTO] <shower>", "[FIND] <water>", "[RINSE] <water>", "[FIND] <soap>", "[WASH] <soap>", "[RINSE] <water>", "[FIND] <towel>", "[GRAB] <towel>", "[SCRUB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'detergent', 'food_butter', 'dough', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'purse', 'faucet', 'stereo', 'picture', 'dustpan', 'bowl', 'dining_room', 'wall', 'oven', 'powersocket', 'photoframe', 'wine', 'tray', 'console', 'pencil', 'sink', 'bathroom_cabinet', 'dishwasher', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'knifeblock', 'drawing', 'cpuscreen', 'clothes_underwear', 'knife', 'floor', 'ceiling', 'cleaning_solution', 'piano_bench', 'toaster', 'ceilinglamp', 'faucet', 'stereo', 'iron', 'wall', 'oven', 'cutting_board', 'tape', 'check', 'coin', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'bag', 'home_office', 'instrument_violin', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Washing hands", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <cleaning_solution>", "[SQUEEZE] <cleaning_solution>", "[SCRUB] <hands_both>", "[SWITCHON] <faucet>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'blender', 'freezer', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'orchid', 'hanger', 'floor', 'bedroom', 'keyboard', 'computer', 'stereo', 'window', 'oil', 'couch', 'pillow', 'wall', 'cup', 'bed', 'food_donut', 'curtain', 'centerpiece', 'console', 'shoes', 'dresser', 'maindoor', 'home_office', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Turn on radio, put on favorite channel, listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['toilet', 'detergent', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'bedroom', 'food_cake', 'stovefan', 'mat', 'faucet', 'window', 'wall', 'oven', 'cup', 'photoframe', 'curtain', 'after_shave', 'doorjamb', 'cards', 'shoes', 'food_steak', 'sink', 'bathroom_cabinet', 'light', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash Hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <faucet>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'freezer', 'trashcan', 'drawing', 'floor', 'phone', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'toaster', 'mat', 'faucet', 'window', 'wall', 'wall_clock', 'powersocket', 'box', 'cup', 'bed', 'video_game_controller', 'television', 'doorjamb', 'console', 'kitchen_counter', 'door', 'light', 'teeth']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk over to the phone and pick it up.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'shoe_rack', 'toilet', 'chair', 'band-aids', 'freezer', 'walllamp', 'cat', 'trashcan', 'folder', 'bookshelf', 'deck_of_cards', 'shower', 'mousepad', 'food_kiwi', 'wall', 'powersocket', 'cup', 'bed', 'clothes_gloves', 'keys', 'television', 'kitchen_counter', 'dishrack', 'controller', 'napkin', 'light', 'soap', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go to the living room then go to the cat and touch it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'toilet', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'phone', 'food_cake', 'computer', 'ceilinglamp', 'shoe-shine_kit', 'iron', 'stereo', 'electric_shaver', 'homework', 'balanceball', 'picture', 'window', 'dining_room', 'powersocket', 'cup', 'photoframe', 'mirror', 'tray', 'pencil', 'shoes', 'conditioner', 'sink', 'creditcard']\nTask:Shave\nresponse:\n{\n    \"thoughts\": Pick up razor.  Push the button to turn it on.  Move it in a linear motion against the beard growth.  Make sure to get the bottom of your chin, your cheeks, and mustache area.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <electric_shaver>', '[FIND] <electric_shaver>', '[GRAB] <electric_shaver>', '[FIND] <mirror>', '[TURNTO] <mirror>', '[SWITCHON] <electric_shaver>', '[SWITCHOFF] <electric_shaver>', '[PUTOBJBACK] <electric_shaver>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'food_salt', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'purse', 'computer', 'ceilinglamp', 'window', 'mousepad', 'food_noodles', 'wall', 'powersocket', 'clothes_gloves', 'food_food', 'toilet_paper', 'television', 'check', 'tray', 'doorjamb', 'cards', 'mouse', 'door', 'light', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, put the groceries in the fridge and close it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'microwave', 'bathroom', 'detergent', 'wallshelf', 'clothes_underwear', 'hanger', 'ceiling', 'bookshelf', 'tvstand', 'faucet', 'deck_of_cards', 'window', 'clothes_socks', 'clothes_pants', 'cup', 'standingmirror', 'photoframe', 'check', 'tray', 'kitchen_counter', 'hair', 'laundry_detergent', 'door', 'bathroom_cabinet', 'tooth_paste', 'clothes_shirt', 'washing_machine', 'coffee', 'shoe_rack']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[DROP] <basket_for_clothes>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[GRAB] <laundry_detergent>', '[PUTOBJBACK] <laundry_detergent>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'microwave', 'jelly', 'freezer', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'purse', 'computer', 'iron', 'window', 'mousepad', 'oil', 'wall', 'powersocket', 'bed', 'keys', 'mouthwash', 'wine', 'doorjamb', 'mouse', 'cupboard', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'beer', 'chair', 'cpuscreen', 'bathroom_counter', 'phone', 'desk', 'cleaning_solution', 'bookmark', 'iron', 'mousepad', 'couch', 'dustpan', 'food_jam', 'tape', 'television', 'wooden_spoon', 'curtain', 'rag', 'doorjamb', 'hairbrush', 'cupboard', 'door', 'bathroom_cabinet', 'table', 'vase', 'colander', 'teeth']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'microwave', 'freezer', 'trashcan', 'toothbrush', 'orchid', 'hanger', 'desk', 'keyboard', 'bookmark', 'shoe-shine_kit', 'iron', 'tvstand', 'plate', 'mousepad', 'microphone', 'wall', 'bench', 'toilet_paper', 'television', 'toothbrush_holder', 'ice', 'shoes', 'chessboard', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'tooth_paste', 'teeth', 'creditcard']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, grab the toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now wash the teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'tablelamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'purse', 'computer', 'clothes_jacket', 'toaster', 'stereo', 'mousepad', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'bed', 'novel', 'mouse', 'shoes', 'dresser', 'clothes_dress', 'light', 'table']\nTask:Get ready for day\nresponse:\n{\n    \"thoughts\": Grab shoes, purse etc and head out the door!,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTON] <clothes_dress>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>', '[FIND] <purse>', '[GRAB] <purse>', '[CLOSE] <dresser>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['glue', 'fork', 'food_pizza', 'tablelamp', 'crayon', 'measuring_cup', 'dough', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'food_cake', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'powersocket', 'bench', 'cup', 'colander', 'mouse', 'hairbrush', 'sink', 'spoon', 'bathroom_cabinet', 'light', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "go to the bathroom and wash your hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[SQUEEZE] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'floor', 'character', 'desk', 'bedroom', 'keyboard', 'thread', 'toaster', 'plate', 'couch', 'wall', 'light_bulb', 'novel', 'coffe_maker', 'food_rice', 'tray', 'food_onion', 'after_shave', 'curtain', 'kettle', 'kitchen_counter', 'shoes', 'cupboard', 'conditioner', 'door', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'water_glass', 'glue', 'walllamp', 'detergent', 'food_carrot', 'toothbrush', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'computer', 'ceilinglamp', 'tvstand', 'window', 'couch', 'clothes_pants', 'form', 'pillow', 'wall', 'nightstand', 'box', 'curtain', 'pencil', 'headset', 'dresser', 'table', 'board_game', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Drink a glass of water in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'toaster', 'stovefan', 'faucet', 'towel', 'pillow', 'wall', 'wall_clock', 'bed', 'check', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'controller', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Cleaning hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'blender', 'freezer', 'walllamp', 'dough', 'orchid', 'drawing', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'window', 'picture', 'couch', 'pillow', 'bowl', 'wall', 'powersocket', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'dishrack', 'door', 'sink', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn on the TV, sit on the couch, find a show, and watch it.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'juice', 'microwave', 'freezer', 'blow_dryer', 'ground_coffee', 'cat', 'floor', 'legs_both', 'face', 'toaster', 'laptop', 'homework', 'closetdrawer', 'towel', 'feet_both', 'arms_both', 'basket_for_clothes', 'newspaper', 'photoframe', 'coffe_maker', 'mouthwash', 'centerpiece', 'doorjamb', 'headset', 'dresser', 'cupboard', 'napkin', 'soap', 'clothes_shirt']\nTask:Try yourself off\nresponse:\n{\n    \"thoughts\": After your shower, go into your bedroom and use a towel to dry water off of yourself before dressing.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <dresser>', '[FIND] <face>', '[WIPE] <face>', '[FIND] <arms_both>', '[WIPE] <arms_both>', '[FIND] <legs_both>', '[WIPE] <legs_both>', '[FIND] <feet_both>', '[WIPE] <feet_both>', '[FIND] <basket_for_clothes>', '[PUTBACK] <towel> <basket_for_clothes>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'plate', 'homework', 'mousepad', 'dining_room', 'wall', 'bench', 'newspaper', 'novel', 'check', 'mouse', 'kitchen_counter', 'hair', 'door', 'sink', 'home_office', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I sit at my computer desk and open the browser.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'freezer', 'walllamp', 'ceiling', 'bedroom', 'keyboard', 'toaster', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'clothes_pants', 'food_orange', 'nightstand', 'bench', 'photoframe', 'bed', 'food_food', 'tray', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'bathtub', 'bathroom_cabinet', 'table', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk into laundry room. Open washing machine door. Put dirty clothes into washing machine. Measure laundry detergent and place into washing machine. Turn dial to appropriate wash setting. Power on.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[CLOSE] <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[OPEN] <washing_machine>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'microwave', 'ground_coffee', 'walllamp', 'detergent', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'toaster', 'stovefan', 'faucet', 'wall', 'pasta', 'oven', 'coffe_maker', 'food_food', 'check', 'coffee_filter', 'television', 'water', 'tray', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'clothes_scarf', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'detergent', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'remote_control', 'button', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'clothes_socks', 'couch', 'wall', 'basket_for_clothes', 'light_bulb', 'television', 'razor', 'curtain', 'food_onion', 'towel_rack', 'coffee_pot', 'pencil', 'kitchen_counter', 'bag', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['scissors', 'blender', 'freezer', 'walllamp', 'trashcan', 'orchid', 'knife', 'bathroom_counter', 'shampoo', 'character', 'desk', 'folder', 'keyboard', 'stovefan', 'mat', 'picture', 'window', 'food_noodles', 'bench', 'toothbrush_holder', 'food_donut', 'after_shave', 'doorjamb', 'cards', 'kitchen_counter', 'vacuum_cleaner', 'dishrack', 'door', 'home_office', 'bathroom_cabinet']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office. I find the telephone. I pick up the telephone handset.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'dog', 'cpuscreen', 'keyboard', 'ceilinglamp', 'plate', 'laptop', 'faucet', 'window', 'stereo', 'oil', 'food_noodles', 'mousepad', 'dining_room', 'nightstand', 'food_food', 'check', 'wine', 'curtain', 'after_shave', 'comb', 'dresser', 'napkin', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I reach out arm. I put hand on radio. I turn on radio. I pick a station.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[TOUCH] <stereo>', '[TURNTO] <stereo>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'chair', 'detergent', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'purse', 'computer', 'toaster', 'mat', 'tvstand', 'clothes_socks', 'food_peanut_butter', 'microphone', 'food_jam', 'pillow', 'wall', 'powersocket', 'after_shave', 'doorjamb', 'cd', 'pajamas', 'laundry_detergent', 'door', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Coming home turning on lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'mousepad', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'microwave', 'walllamp', 'trashcan', 'knife', 'drawing', 'floor', 'computer', 'toaster', 'mat', 'plate', 'faucet', 'ceilinglamp', 'stovefan', 'dish_soap', 'oil', 'wall', 'powersocket', 'tape', 'coffe_maker', 'television', 'drinking_glass', 'rag', 'shaving_cream', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'clothes_skirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and turn to the dishwasher, put the knife, glass, fork and plate into the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[TURNTO] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'newspaper', 'television', 'doorjamb', 'dresser', 'soap', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit on my couch and watch TV in my living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'alcohol', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'iron', 'faucet', 'mousepad', 'bowl', 'wall', 'nightstand', 'oven', 'mouse', 'cupboard', 'door', 'clothes_scarf', 'bathroom_cabinet', 'table', 'instrument_guitar']\nTask:Work\nresponse:\n", "output": {"thoughts": "For doing homework walk to home office. Walk to computer. Switch on Computer. Sit in a chair and do homework in computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[WALK] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'food_salt', 'beer', 'walllamp', 'drawing', 'orchid', 'knifeblock', 'clothes_underwear', 'floor', 'folder', 'man', 'computer', 'toaster', 'ceilinglamp', 'stereo', 'window', 'homework', 'mousepad', 'couch', 'microphone', 'pillow', 'wall', 'nightstand', 'keys', 'food_food', 'wooden_spoon', 'curtain', 'doorjamb', 'hair', 'table']\nTask:Have conversation with boyfriend\nresponse:\n{\n    \"thoughts\": Sit on sofa, talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'deck_of_cards', 'electric_shaver', 'mousepad', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'newspaper', 'bed', 'food_food', 'wine', 'doorjamb', 'cards', 'mouse', 'shoes', 'dresser', 'door', 'light', 'dirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'woman', 'knifeblock', 'needle', 'cpuscreen', 'desk', 'man', 'keyboard', 'bedroom', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'nightstand', 'oven', 'cup', 'food_food', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'ground_coffee', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'food_cake', 'mat', 'tvstand', 'shoe-shine_kit', 'homework', 'oil', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'bench', 'oven', 'powersocket', 'toilet_paper', 'kitchen_counter', 'door', 'tooth_paste']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk over and pick up the phone", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'freezer', 'knife', 'floor', 'ceiling', 'folder', 'stamp', 'brush', 'mat', 'ceilinglamp', 'shower', 'faucet', 'toaster', 'pillow', 'bowl', 'food_kiwi', 'wall', 'powersocket', 'toilet_paper', 'mouthwash', 'doorjamb', 'door', 'home_office', 'light', 'table']\nTask:Turn light off\nresponse:\n{\n    \"thoughts\": Put the light switch into the down, or off, position.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['fork', 'glue', 'board_game', 'toilet', 'bathroom', 'dough', 'toothbrush', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'mop_bucket', 'mat', 'faucet', 'window', 'wall', 'photoframe', 'food_food', 'toothbrush_holder', 'razor', 'tray', 'after_shave', 'doorjamb', 'vacuum_cleaner', 'conditioner', 'sink', 'napkin', 'bathroom_cabinet', 'soap', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I go into the bathroom, go to the sink, get toothbrush and toothpaste. Put toothpaste on the toothbrush, brush teeth, turn on faucet, rinse brush, and return toothbrush.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'food_pizza', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'cleaning_solution', 'mail', 'brush', 'toaster', 'tvstand', 'balanceball', 'window', 'food_noodles', 'mousepad', 'food_peanut_butter', 'clothes_hat', 'dining_room', 'pasta', 'bench', 'photoframe', 'oven_mitts', 'ice', 'rag', 'dresser', 'cupboard', 'food_steak', 'maindoor', 'table']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'walllamp', 'trashcan', 'dough', 'toothbrush', 'floor', 'dvd_player', 'bedroom', 'stovefan', 'mat', 'tvstand', 'window', 'food_noodles', 'dining_room', 'nightstand', 'powersocket', 'oven', 'cutting_board', 'cup', 'light_bulb', 'coffe_maker', 'mouthwash', 'oven_mitts', 'doorjamb', 'diary', 'bag', 'home_office', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "go to living room, find phone, pick up phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'ground_coffee', 'knifeblock', 'cpuscreen', 'longboard', 'floor', 'phone', 'bedroom', 'toaster', 'faucet', 'pillow', 'wall', 'bed', 'novel', 'water', 'tray', 'centerpiece', 'cards', 'kitchen_counter', 'shoes', 'dresser', 'cupboard', 'conditioner', 'sink', 'bag', 'napkin', 'table', 'dishwasher', 'board_game']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <sink>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'jelly', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'food_chicken', 'desk', 'keyboard', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'stovefan', 'faucet', 'mousepad', 'oil', 'wall', 'powersocket', 'oven', 'food_food', 'tray', 'doorjamb', 'coffee_pot', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'floor_lamp', 'mousepad', 'couch', 'dining_room', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'doorjamb', 'cd', 'door', 'bathtub', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to wall, find light switch, turn on lights,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[WALK] <bedroom>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'dry_pasta', 'bathroom', 'toothbrush', 'drawing', 'knifeblock', 'phone', 'character', 'clothes_jacket', 'piano_bench', 'mat', 'shoe-shine_kit', 'stereo', 'window', 'oil', 'nightstand', 'cup', 'clothes_gloves', 'check', 'wine', 'tray', 'console', 'pencil', 'dresser', 'hair', 'door', 'clothes_skirt', 'dishwasher', 'clothes_shirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Pet Cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'food_butter', 'drawing', 'bills', 'cpuscreen', 'longboard', 'floor', 'bedroom', 'desk', 'bookshelf', 'button', 'bookmark', 'toaster', 'ceilinglamp', 'deck_of_cards', 'iron', 'tvstand', 'pillow', 'pasta', 'bench', 'powersocket', 'bed', 'coffe_maker', 'television', 'ice', 'mouse', 'dresser', 'controller']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk into living room, look at the phone when it ringing, grab the phone or pickup the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>']\n}\n", "input": "\nScene_Objects:['walllamp', 'toothbrush', 'drawing', 'orchid', 'floor', 'ceiling', 'remote_control', 'clothes_jacket', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'oven', 'box', 'basket_for_clothes', 'novel', 'check', 'television', 'tray', 'curtain', 'doorjamb', 'door', 'table', 'clothes_skirt', 'creditcard']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to living room. Walk to and sit at the sofa. Grab remote control and switch on TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'pillow', 'wall', 'nightstand', 'photoframe', 'wine', 'doorjamb', 'mouse', 'dresser', 'napkin', 'bathroom_cabinet', 'light']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Wipe down computer desk and monitor,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <desk>', '[FIND] <computer>', '[WIPE] <computer>', '[WIPE] <desk>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'trashcan', 'drawing', 'floor', 'character', 'toaster', 'plate', 'tvstand', 'foundation', 'dish_soap', 'tooth_paste', 'wall', 'cup', 'coffe_maker', 'food_food', 'food_onion', 'centerpiece', 'cards', 'kettle', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'door', 'sink', 'light', 'dishwasher', 'board_game', 'clothes_shirt', 'creditcard']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Wash a plate using the dishwasher in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'sheets', 'pot', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mail', 'tvstand', 'stereo', 'window', 'mousepad', 'clothes_socks', 'wall', 'nightstand', 'powersocket', 'pasta', 'oven', 'light_bulb', 'coffe_maker', 'food_rice', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'door', 'light']\nTask:Turn light off\nresponse:\n{\n    \"thoughts\": Put the light switch into the down, or off, position.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['walllamp', 'trashcan', 'toothbrush', 'drawing', 'orchid', 'needle', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'bookshelf', 'remote_control', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'wall', 'powersocket', 'television', 'food_rice', 'doorjamb', 'pencil', 'chessboard', 'door', 'hairdryer', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'chair', 'jelly', 'trashcan', 'toothbrush', 'orchid', 'cpuscreen', 'character', 'bookshelf', 'button', 'mat', 'deck_of_cards', 'faucet', 'shower', 'pillow', 'powersocket', 'bench', 'photoframe', 'bed', 'novel', 'coin', 'kitchen_counter', 'headset', 'chessboard', 'comb', 'dresser', 'table', 'coffee', 'creditcard']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to the entrance hall. Find the telephone. Grab the telephone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'jelly', 'trashcan', 'toothbrush', 'orchid', 'cpuscreen', 'character', 'bookshelf', 'button', 'mat', 'deck_of_cards', 'faucet', 'shower', 'pillow', 'powersocket', 'bench', 'photoframe', 'bed', 'novel', 'coin', 'kitchen_counter', 'headset', 'chessboard', 'comb', 'dresser', 'table', 'coffee', 'creditcard']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the entrance hall. Find the telephone. Grab the telephone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'tablelamp', 'toilet', 'jelly', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'coffee_filter', 'curtain', 'doorjamb', 'centerpiece', 'door', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit down on sofa, get the remote, turn on tv, change channel to show I want to watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['freezer', 'cat', 'milk', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'towel', 'wall', 'powersocket', 'oven', 'food_food', 'coffe_maker', 'coin', 'doorjamb', 'mouse', 'headset', 'shoes', 'pajamas', 'door', 'light', 'colander', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk to the kitchen, open the fridge and put the groceries in the fridge. Then I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'jelly', 'detergent', 'food_carrot', 'toothbrush', 'needle', 'hanger', 'floor', 'phone', 'shampoo', 'character', 'tvstand', 'iron', 'balanceball', 'stereo', 'mousepad', 'food_jam', 'pillow', 'nightstand', 'toothbrush_holder', 'diary', 'console', 'shaving_cream', 'kitchen_counter', 'sink', 'light', 'dishwasher', 'vase']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the playback device you want to listen to. turn on. press play. enjoyu music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'freezer', 'bathroom', 'knifeblock', 'hanger', 'character', 'desk', 'bedroom', 'folder', 'toaster', 'mat', 'laptop', 'stereo', 'foundation', 'homework', 'bed', 'newspaper', 'television', 'water', 'wine', 'tray', 'kitchen_counter', 'chessboard', 'laundry_detergent', 'food_steak', 'bag', 'home_office', 'light', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "put soap on hands, scrub it and rinse with water", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'ceilinglamp', 'faucet', 'shower', 'towel', 'dining_room', 'wall', 'powersocket', 'bed', 'check', 'wooden_spoon', 'cards', 'mouse', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom. turn on the water in the sink. apply soap to hands. rub hands together. rinse hands with water. turn off the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_pizza', 'scissors', 'band-aids', 'nail_polish', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'dvd_player', 'bookshelf', 'thread', 'computer', 'button', 'stereo', 'window', 'clothes_socks', 'couch', 'food_peanut_butter', 'dining_room', 'toilet_paper', 'tray', 'mouse', 'dishrack', 'comb', 'bag', 'home_office', 'clothes_shirt', 'shoe_rack']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter living room. Walk to telephone. Pick up from hook.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_bread', 'microwave', 'walllamp', 'ceilingfan', 'food_butter', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'food_orange', 'wall', 'bench', 'photoframe', 'keys', 'check', 'food_food', 'television', 'curtain', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Go to the living room, Switch on the television. Find the remote and press the change channel button, change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[TOUCH] <button>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'tea', 'toilet', 'walllamp', 'drawing', 'orchid', 'floor', 'desk', 'bedroom', 'computer', 'ceilinglamp', 'tvstand', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'food_cheese', 'bed', 'food_donut', 'curtain', 'food_onion', 'mouse', 'pajamas', 'controller', 'table', 'dishwasher', 'light', 'board_game']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "lie in the sofa, relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'microwave', 'toilet', 'jelly', 'freezer', 'band-aids', 'drawing', 'knifeblock', 'clothes_underwear', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'powersocket', 'cup', 'basket_for_clothes', 'photoframe', 'food_food', 'water', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Get drink\nresponse:\n{\n    \"thoughts\": I go to the fridge and open the door. I look for the water pitcher, and take it out, making sure to close the fridge door behind me. Then I take a pint glass out of the upper cabinet over the sink and use the water pitcher to pour a full glass of water. Then I return the pitcher to the fridge and shut the door again, and take my glass of water out of the kitche.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[OPEN] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'toilet', 'chair', 'walllamp', 'bathroom', 'food_butter', 'drawing', 'orchid', 'hanger', 'floor', 'desk', 'tea_bag', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'window', 'couch', 'candle', 'food_orange', 'pillow', 'wall', 'cup', 'bed', 'curtain', 'kettle', 'pencil', 'pajamas', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Sit on sofa. Lie down on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'freezer', 'knifeblock', 'knife', 'bathroom_counter', 'shampoo', 'ceiling', 'keyboard', 'stovefan', 'mat', 'stereo', 'clothes_socks', 'oil', 'nightstand', 'powersocket', 'toy', 'novel', 'food_food', 'coffe_maker', 'rag', 'kettle', 'shoes', 'cupboard', 'hair', 'controller', 'door', 'instrument_guitar', 'table', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk into living room, walk to stereo, turn on stereo, listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['crayon', 'mop', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'alcohol', 'ceilinglamp', 'mat', 'faucet', 'mousepad', 'clothes_hat', 'pillow', 'wall', 'bed', 'newspaper', 'food_food', 'coffee_filter', 'mouse', 'pencil', 'door', 'sink', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go to home office and browse the internet.", "actions": ["[RUN] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TOUCH] <mouse>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TOUCH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'toilet', 'chair', 'ground_coffee', 'bathroom', 'dough', 'wallshelf', 'phone', 'character', 'keyboard', 'computer', 'ceilinglamp', 'deck_of_cards', 'stereo', 'mousepad', 'couch', 'powersocket', 'cutting_board', 'bed', 'food_food', 'food_rice', 'tray', 'wooden_spoon', 'pencil', 'headset', 'cupboard', 'laundry_detergent', 'sink', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find the cat, go near the cat and pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'dog', 'cat', 'drawing', 'food_egg', 'floor', 'phone', 'piano_bench', 'mat', 'mousepad', 'oil', 'food_jam', 'pillow', 'wall_clock', 'photoframe', 'coffe_maker', 'food_donut', 'food_rice', 'tray', 'after_shave', 'kitchen_counter', 'clothes_dress', 'door', 'sink', 'bag', 'bathroom_cabinet', 'table', 'dishwasher', 'creditcard']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room , find phone ,pick up phone 3", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'microwave', 'walllamp', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'bedroom', 'ceilinglamp', 'plate', 'faucet', 'window', 'dish_soap', 'mousepad', 'bowl', 'wall', 'oven', 'powersocket', 'check', 'mouthwash', 'after_shave', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": I go to the kitchen and turn on the faucet. I get dish soap and put it into the sink. I grab a sponge and a bowl and start cleaning it. I put the bowl into the dishrack. I find a plate and clean it as well. I put it into the dishrack, turn off the faucet and put the sponge back.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <bowl>', '[GRAB] <bowl>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <bowl>', '[SCRUB] <bowl>', '[RINSE] <bowl>', '[FIND] <dishrack>', '[PUTBACK] <bowl> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <sponge>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'toilet', 'nail_polish', 'detergent', 'cat', 'trashcan', 'milk', 'knifeblock', 'phone', 'ceiling', 'desk', 'keyboard', 'mousepad', 'couch', 'pillow', 'bench', 'oven', 'cup', 'bed', 'light_bulb', 'food_food', 'lighter', 'kitchen_counter', 'hairbrush', 'dresser', 'laundry_detergent', 'table', 'clothes_skirt', 'tooth_paste', 'dirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting a cat may sound simple.Letting the Cat Come to You.Extend a hand or finger and allow the cat a chance to touch her nose to you. When a cat bumps her head into your hand, it's a signal she wants attention.Focusing on Areas with Scent Glands. Feed with food.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <milk>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[TURNTO] <cat>", "[POINTAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'toilet', 'walllamp', 'detergent', 'food_carrot', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'bedroom', 'keyboard', 'bookshelf', 'mat', 'mousepad', 'dining_room', 'wall', 'nightstand', 'bench', 'basket_for_clothes', 'bed', 'photoframe', 'coffe_maker', 'food_food', 'wooden_spoon', 'mouse', 'dresser', 'sink', 'dishwasher']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go t the bedroom and go to sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['sheets', 'water_glass', 'pot', 'spectacles', 'chair', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'ceiling', 'character', 'stovefan', 'mat', 'faucet', 'food_orange', 'wall', 'oven', 'cutting_board', 'food_food', 'water', 'tray', 'food_onion', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'instrument_guitar', 'home_office']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen, take a glass from the cupboard, fill with water and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>", "[RINSE] <water_glass>", "[PUTBACK] <water_glass> <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'microwave', 'toilet', 'cat', 'wallshelf', 'bathroom_counter', 'floor', 'dvd_player', 'bookshelf', 'stovefan', 'ceilinglamp', 'toaster', 'mat', 'shower', 'faucet', 'tvstand', 'couch', 'towel', 'wall', 'bench', 'cutting_board', 'bed', 'razor', 'television', 'curtain', 'kitchen_counter', 'dresser', 'home_office', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Pet Cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'filing_cabinet', 'tablelamp', 'beer', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'cat', 'cpuscreen', 'phone', 'character', 'food_cake', 'computer', 'piano_bench', 'brush', 'clothes_pants', 'dining_room', 'nightstand', 'cutting_board', 'toilet_paper', 'coffe_maker', 'check', 'curtain', 'kitchen_counter', 'table', 'scrabble', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Every time this animal is seen, you must approach it and show compassion for it by petting it in a gentle manner.", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <bedroom>", "[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[FIND] <brush>", "[GRAB] <brush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'beer', 'microwave', 'chair', 'drawing', 'hanger', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'bookmark', 'laptop', 'ceilinglamp', 'faucet', 'shower', 'mousepad', 'couch', 'food_peanut_butter', 'photoframe', 'newspaper', 'razor', 'tray', 'kitchen_counter', 'vacuum_cleaner', 'clothes_dress', 'door', 'bathroom_cabinet', 'tooth_paste']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I go to my bed room and sit on the chair and in front of me is my desktop computer and start writing an email.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'blender', 'crayon', 'microwave', 'toilet', 'walllamp', 'milk', 'drawing', 'bathroom_counter', 'phone', 'desk', 'envelope', 'table_cloth', 'ceilinglamp', 'plate', 'faucet', 'picture', 'couch', 'clothes_pants', 'powersocket', 'bench', 'newspaper', 'tray', 'rag', 'kitchen_counter', 'laundry_detergent', 'soap', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[PLUGIN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'character', 'bedroom', 'desk', 'bookshelf', 'computer', 'mat', 'tvstand', 'window', 'pillow', 'wall', 'powersocket', 'tray', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'bathtub', 'light', 'teeth', 'shoe_rack']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['blender', 'band-aids', 'freezer', 'walllamp', 'nail_polish', 'trashcan', 'dough', 'drawing', 'bills', 'cpuscreen', 'bathroom_counter', 'ceiling', 'phone', 'character', 'desk', 'purse', 'alcohol', 'clothes_jacket', 'toaster', 'mat', 'faucet', 'oil', 'couch', 'broom', 'oven', 'photoframe', 'tray', 'kitchen_counter', 'dresser', 'laser_pointer']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find a cat in the kitchen and touch it", "actions": ["[WALK] <dining_room>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sponge', 'food_bread', 'blender', 'chair', 'walllamp', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'tvstand', 'window', 'mousepad', 'clothes_socks', 'wall', 'oven', 'cup', 'television', 'mouse', 'cupboard', 'door', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": go to the office, turn computer on, open an browser or any application to open any website.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <computer>']\n}\n", "input": "\nScene_Objects:['food_salt', 'tea', 'microwave', 'freezer', 'ground_coffee', 'walllamp', 'bathroom', 'trashcan', 'floor', 'ceiling', 'mop_bucket', 'toaster', 'picture', 'electric_shaver', 'pillow', 'wall', 'coffe_maker', 'food_food', 'coffee_filter', 'oven_mitts', 'water', 'ice', 'coffee_pot', 'kitchen_counter', 'cd', 'dishrack', 'cupboard', 'instrument_guitar', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making Coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <water>", "[GRAB] <water>", "[OPEN] <coffe_maker>", "[POUR] <water> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <coffee_pot>", "[GRAB] <coffee_pot>", "[PUTBACK] <coffee_pot> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'curtain', 'shoes', 'cd', 'vacuum_cleaner', 'dresser', 'cupboard', 'clothes_dress', 'home_office', 'table', 'shoe_rack']\nTask:Take off shoes\nresponse:\n{\n    \"thoughts\": Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['scissors', 'food_ice_cream', 'walllamp', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'shampoo', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'check', 'television', 'coin', 'curtain', 'doorjamb', 'cards', 'shoes', 'cupboard', 'door', 'sink', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to the living room to watch TV", "actions": ["[RUN] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[TOUCH] <remote_control>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[DROP] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'blender', 'toilet', 'jelly', 'chair', 'drawing', 'bathroom_counter', 'floor', 'character', 'food_bacon', 'mat', 'shoe-shine_kit', 'shower', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'pasta', 'oven', 'newspaper', 'check', 'television', 'ice', 'tray', 'doorjamb', 'pencil', 'comb', 'dresser', 'bathroom_cabinet']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom. Pull down pants. Use the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <clothes_pants>', '[FIND] <clothes_pants>', '[PULL] <clothes_pants>', '[FIND] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'juice', 'walllamp', 'bathroom', 'drawing', 'hanger', 'floor', 'character', 'desk', 'tea_bag', 'mop_bucket', 'plate', 'faucet', 'couch', 'dining_room', 'broom', 'nightstand', 'bench', 'powersocket', 'box', 'photoframe', 'coffe_maker', 'food_food', 'television', 'dresser', 'cupboard', 'home_office', 'tooth_paste', 'coffee']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the living room. I find the stereo and turn it on, and I listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'freezer', 'food_egg', 'cpuscreen', 'floor', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'laptop', 'picture', 'mousepad', 'dining_room', 'wall', 'nightstand', 'photoframe', 'bed', 'mouse', 'kitchen_counter', 'maindoor', 'home_office', 'light', 'clothes_skirt']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": walk to home office and switch on laptop and with the useof mouse and keyboard send mail,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'tea', 'band-aids', 'walllamp', 'nail_polish', 'knifeblock', 'bills', 'cpuscreen', 'hanger', 'shampoo', 'phone', 'floor', 'bedroom', 'desk', 'toaster', 'stereo', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bowl', 'dining_room', 'bench', 'newspaper', 'clothes_gloves', 'coffe_maker', 'curtain', 'diary', 'hairbrush', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'chair', 'freezer', 'cat', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'stereo', 'faucet', 'picture', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'cup', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'home_office', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to the bedroom and sit at my desk. I turn my radio on and listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['fork', 'juice', 'food_ice_cream', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'faucet', 'food_noodles', 'mousepad', 'clothes_socks', 'wall', 'nightstand', 'box', 'photoframe', 'coffe_maker', 'mouse', 'kitchen_counter', 'door', 'food_steak', 'table', 'colander', 'creditcard']\nTask:Work\nresponse:\n", "output": {"thoughts": "turn computer on, open application to work with", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'foundation', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'cards', 'door', 'table', 'light', 'coffee']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the sofa to watch tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tea', 'microwave', 'toilet', 'walllamp', 'detergent', 'trashcan', 'floor', 'ceiling', 'envelope', 'toaster', 'plate', 'shower', 'stereo', 'dish_soap', 'wall', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'food_food', 'mouthwash', 'check', 'curtain', 'kitchen_counter', 'cupboard', 'pajamas', 'conditioner', 'sink', 'spoon', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'food_salt', 'toilet', 'walllamp', 'bathroom', 'food_butter', 'drawing', 'hanger', 'shampoo', 'bedroom', 'purse', 'computer', 'stovefan', 'laptop', 'ceilinglamp', 'electric_shaver', 'window', 'homework', 'clothes_socks', 'wall_clock', 'shelf', 'basket_for_clothes', 'bed', 'coffe_maker', 'check', 'wooden_spoon', 'centerpiece', 'lighter', 'kettle']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['food_bread', 'microwave', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'keyboard', 'computer', 'alcohol', 'toaster', 'stovefan', 'shower', 'dustpan', 'pillow', 'wall', 'nightstand', 'bench', 'coffe_maker', 'check', 'food_food', 'coffee_filter', 'water', 'kitchen_counter', 'cupboard', 'table', 'dishwasher', 'tooth_paste', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Go to the kitchen and make coffee.", "actions": ["[RUN] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[PUTOBJBACK] <coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'detergent', 'food_butter', 'dough', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'purse', 'faucet', 'stereo', 'picture', 'dustpan', 'bowl', 'dining_room', 'wall', 'oven', 'powersocket', 'photoframe', 'wine', 'tray', 'console', 'pencil', 'sink', 'bathroom_cabinet', 'dishwasher', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'walllamp', 'drawing', 'orchid', 'floor', 'phone', 'keyboard', 'brush', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'stereo', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'curtain', 'cd', 'dishrack', 'dresser', 'cupboard', 'table', 'laser_pointer', 'clothes_shirt', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room,find the sofa,sat down in the sofa,take some relaxation.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'trashcan', 'dough', 'food_carrot', 'knife', 'drawing', 'clothes_underwear', 'hanger', 'dvd_player', 'character', 'bookmark', 'toaster', 'mat', 'faucet', 'oven', 'tape', 'food_food', 'television', 'coffee_filter', 'ice', 'wooden_spoon', 'tray', 'lighter', 'chessboard', 'spoon', 'home_office', 'table', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I hear the phone ringing in the living room. I walk to the living room and go to the phone. I pick it up and answer the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>', '[SWITCHON] <phone>']\n}\n", "input": "\nScene_Objects:['fork', 'food_salt', 'toilet', 'bathroom', 'nail_polish', 'toothbrush', 'phone', 'character', 'mat', 'shower', 'stereo', 'towel', 'dining_room', 'oven', 'basket_for_clothes', 'novel', 'television', 'oven_mitts', 'food_donut', 'tray', 'doorjamb', 'cards', 'kitchen_counter', 'dishrack', 'pajamas', 'door', 'napkin', 'light', 'table', 'soap']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room, sit down by the cat, put out hand, stroke the cat from head to tail", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'mop', 'microwave', 'ground_coffee', 'walllamp', 'cat', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'toaster', 'brush', 'tvstand', 'shower', 'faucet', 'food_jam', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'mouthwash', 'coffee_filter', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'dishwasher', 'scrabble']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Make coffee in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[PLUGIN] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'microwave', 'walllamp', 'bathroom', 'cat', 'trashcan', 'knifeblock', 'hanger', 'floor', 'faucet', 'shower', 'clothes_hat', 'wall', 'nightstand', 'tape', 'cup', 'food_food', 'water', 'tray', 'diary', 'pencil', 'kitchen_counter', 'headset', 'hairbrush', 'cupboard', 'hair', 'sink', 'instrument_guitar', 'soap']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'freezer', 'bathroom', 'orchid', 'phone', 'dvd_player', 'character', 'bedroom', 'ceilinglamp', 'balanceball', 'iron', 'window', 'couch', 'food_orange', 'wall', 'tape', 'cup', 'basket_for_clothes', 'check', 'wooden_spoon', 'console', 'mouse', 'kitchen_counter', 'hairbrush', 'food_steak', 'sink', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'bathroom_counter', 'floor', 'mop_bucket', 'table_cloth', 'toaster', 'stovefan', 'plate', 'microphone', 'clothes_pants', 'dining_room', 'wall', 'pasta', 'cup', 'coffe_maker', 'food_food', 'mouse', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'home_office', 'light', 'dishwasher', 'colander', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'microwave', 'chair', 'freezer', 'walllamp', 'detergent', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'stereo', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door', 'bag', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['chair', 'jelly', 'walllamp', 'dough', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'stamp', 'keyboard', 'food_cake', 'computer', 'mop_bucket', 'mat', 'tvstand', 'mousepad', 'dining_room', 'wall', 'bench', 'toy', 'food_cheese', 'novel', 'curtain', 'mouse', 'shoes', 'sink', 'table', 'creditcard']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I walk to home office and search for a book, and sit to the chair with a table, and start reading.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'picture', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <chair>', '[SIT] <chair>', '[TURNTO] <television>', '[WATCH] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'crayon', 'dog', 'freezer', 'knife', 'orchid', 'bathroom_counter', 'character', 'bedroom', 'faucet', 'couch', 'food_orange', 'towel', 'nightstand', 'oven', 'powersocket', 'photoframe', 'check', 'video_game_controller', 'food_donut', 'food_rice', 'food_onion', 'diary', 'dresser', 'hairbrush', 'clothes_scarf', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the cat. Pet the cat with your hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'dog', 'freezer', 'drawing', 'cpuscreen', 'phone', 'tea_bag', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'window', 'oil', 'food_orange', 'food_cheese', 'clothes_gloves', 'toilet_paper', 'food_food', 'television', 'coffee_filter', 'check', 'curtain', 'kettle', 'mouse', 'food_steak', 'spoon', 'instrument_guitar', 'light', 'dishwasher']\nTask:Walk to room\nresponse:\n{\n    \"thoughts\": Leave the living room and enter the hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['pot', 'water_glass', 'dog', 'walllamp', 'knifeblock', 'bills', 'floor', 'ceiling', 'stovefan', 'brush', 'mat', 'faucet', 'homework', 'wall', 'pasta', 'oven', 'powersocket', 'food_food', 'water', 'tray', 'coffee_pot', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'board_game', 'clothes_shirt', 'dirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'floor', 'bedroom', 'bookshelf', 'keyboard', 'computer', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'novel', 'check', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'light', 'table']\nTask:Read\nresponse:\n{\n    \"thoughts\": Read a book in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'toothbrush', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'remote_control', 'clothes_jacket', 'ceilinglamp', 'tvstand', 'mat', 'picture', 'window', 'faucet', 'couch', 'clothes_hat', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go into the living room. Walk over to the sofa. Lie down on the sofa. Turn on and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'juice', 'microwave', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'shower', 'faucet', 'window', 'mousepad', 'food_orange', 'wall', 'television', 'curtain', 'after_shave', 'mouse', 'dresser', 'door', 'napkin', 'table']\nTask:Clean screen\nresponse:\n{\n    \"thoughts\": Use a soft rag to wipe the screen until there are no marks on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <computer>', '[WIPE] <computer>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'walllamp', 'trashcan', 'drawing', 'floor', 'ceiling', 'button', 'toaster', 'window', 'mousepad', 'clothes_pants', 'wall', 'basket_for_clothes', 'coffe_maker', 'check', 'television', 'toothbrush_holder', 'ice', 'diary', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'pajamas', 'napkin', 'coffee_cup', 'dishwasher', 'soap', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to the kitchen at the coffee maker, grab coffee, pour coffee into coffee maker, put coffee into coffee maker, push coffee maker's button, wait to complete, grab coffee cup", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[OPEN] <coffe_maker>", "[POUR] <coffee> <coffe_maker>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'keyboard', 'mat', 'ceilinglamp', 'shower', 'oil', 'food_peanut_butter', 'food_orange', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'pencil', 'dresser', 'door', 'napkin', 'light', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": put away groceries in fridge. open the door, put on shelf, close the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'mat', 'faucet', 'couch', 'dining_room', 'wall', 'newspaper', 'toilet_paper', 'television', 'oven_mitts', 'water', 'tray', 'vacuum_cleaner', 'controller', 'door', 'sink', 'napkin', 'bathroom_cabinet', 'instrument_guitar', 'light', 'soap', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Wash the teeth", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[TURNTO] <water>", "[POINTAT] <water>", "[FIND] <teeth>", "[WASH] <teeth>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'tea', 'toilet', 'chair', 'freezer', 'drawing', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'dvd_player', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'food_jam', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'cup', 'food_vegetable', 'photoframe', 'basket_for_clothes', 'doorjamb', 'mouse', 'door', 'table', 'light']\nTask:Chop vegetables\nresponse:\n{\n    \"thoughts\": Cut up veggies with knife,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[FIND] <cutting_board>', '[GRAB] <cutting_board>', '[CLOSE] <kitchen_cabinet>', '[WALK] <table>', '[PUTBACK] <knife> <table>', '[PUTBACK] <cutting_board> <table>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <food_vegetable> <cutting_board>', '[GRAB] <knife>', '[CUT] <food_vegetable>', '[PUTOBJBACK] <knife>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'ground_coffee', 'walllamp', 'bathroom', 'trashcan', 'floor', 'ceiling', 'phone', 'computer', 'toaster', 'stovefan', 'homework', 'wall', 'novel', 'coffe_maker', 'television', 'coffee_filter', 'water', 'tray', 'after_shave', 'diary', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'food_steak', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'mat', 'plate', 'faucet', 'window', 'dish_soap', 'food_noodles', 'closetdrawer', 'couch', 'towel', 'dining_room', 'wall', 'wine', 'rag', 'shaving_cream', 'kitchen_counter', 'chessboard', 'dresser', 'cupboard', 'conditioner', 'sink', 'table', 'dishwasher', 'soap']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Run warm water in the sink. Pour dish washing liquid on the dishes. Pick up dishes and wash them individually with a sponge. Rinse suds off under running water. Dry dishes with towel.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[WASH] <plate>', '[RINSE] <plate>', '[RINSE] <sponge>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <sponge>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <plate>', '[PUTOBJBACK] <towel>', '[WALK] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'dresser', 'hairbrush', 'clothes_dress', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'microwave', 'walllamp', 'food_carrot', 'character', 'desk', 'keyboard', 'shower', 'washing_machine', 'window', 'mousepad', 'clothes_pants', 'towel', 'wall', 'powersocket', 'basket_for_clothes', 'check', 'mouthwash', 'food_rice', 'headset', 'chessboard', 'hair', 'napkin', 'table', 'soap', 'scrabble', 'dirt', 'coffee', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[WALK] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'remote_control', 'thread', 'brush', 'ceilinglamp', 'tvstand', 'mat', 'toaster', 'clothes_hat', 'dining_room', 'wall', 'powersocket', 'cup', 'photoframe', 'television', 'doorjamb', 'pencil', 'dresser', 'door', 'home_office', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Change the channel on the TV in the living room", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'microwave', 'freezer', 'ground_coffee', 'dough', 'orchid', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'tea_bag', 'toaster', 'faucet', 'shower', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'curtain', 'kitchen_counter', 'cupboard', 'sink']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'shoe-shine_kit', 'mousepad', 'pillow', 'wall', 'powersocket', 'cup', 'bed', 'wooden_spoon', 'coffee_pot', 'mouse', 'kitchen_counter', 'door', 'home_office', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I walk to the Home Office then find the computer and sit in the chair. I switch on the computer and start typing the email. I switch the computer off.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'freezer', 'walllamp', 'cpuscreen', 'wallshelf', 'ceiling', 'computer', 'shower', 'window', 'mousepad', 'couch', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'bench', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'razor', 'mouse', 'cupboard', 'laundry_detergent', 'door', 'home_office', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'blender', 'food_ice_cream', 'microwave', 'chair', 'walllamp', 'bathroom', 'trashcan', 'milk', 'floor', 'bedroom', 'computer', 'toaster', 'plate', 'shoe-shine_kit', 'oil', 'wall', 'newspaper', 'coffe_maker', 'tray', 'centerpiece', 'console', 'pencil', 'kitchen_counter', 'cupboard', 'spoon', 'dishwasher', 'laser_pointer', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'freezer', 'detergent', 'floor', 'phone', 'ceiling', 'bedroom', 'keyboard', 'stovefan', 'shower', 'faucet', 'window', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'bench', 'photoframe', 'coffe_maker', 'check', 'television', 'food_food', 'ice', 'rag', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'toilet', 'food_carrot', 'toothbrush', 'drawing', 'cpuscreen', 'floor', 'bedroom', 'bookshelf', 'thread', 'tvstand', 'shower', 'clothes_pants', 'towel', 'wall', 'pasta', 'bench', 'photoframe', 'light_bulb', 'food_food', 'television', 'curtain', 'doorjamb', 'kettle', 'door', 'sink', 'hairdryer', 'light']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom. Pull down pants. Use the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[PULL] <clothes_pants>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'desk', 'brush', 'mat', 'ceilinglamp', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'wall', 'powersocket', 'cup', 'photoframe', 'coffee_filter', 'food_rice', 'doorjamb', 'towel_rack', 'dresser', 'clothes_dress', 'door', 'light', 'table']\nTask:Change clothes\nresponse:\n{\n    \"thoughts\": Switch your clothes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTON] <clothes_dress>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['blender', 'beer', 'freezer', 'trashcan', 'orchid', 'floor', 'ceiling', 'bedroom', 'computer', 'toaster', 'window', 'couch', 'candle', 'dining_room', 'broom', 'oven', 'toy', 'check', 'video_game_controller', 'tray', 'kettle', 'coffee_pot', 'mouse', 'cd', 'cupboard', 'controller', 'bathroom_cabinet', 'light', 'dishwasher', 'laser_pointer']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I pick up my phone. I touch it to navigate to the web browser. I can swipe and tap the screen to visit different websites. Then I read or watch the content within those websites.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[TOUCH] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'jelly', 'freezer', 'walllamp', 'drawing', 'character', 'tea_bag', 'stovefan', 'mat', 'tvstand', 'stereo', 'shoe-shine_kit', 'window', 'ceilinglamp', 'dustpan', 'food_jam', 'pillow', 'wall', 'wall_clock', 'newspaper', 'coffe_maker', 'food_donut', 'doorjamb', 'coffee_pot', 'cards', 'door', 'bathroom_cabinet', 'colander', 'scrabble']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I reach out arm. I put hand on radio. I turn on radio. I pick a station.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[TOUCH] <stereo>', '[TURNTO] <stereo>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sheets', 'toilet', 'chair', 'freezer', 'ground_coffee', 'bathroom', 'cat', 'milk', 'food_egg', 'phone', 'clothes_jacket', 'mat', 'plate', 'tvstand', 'stereo', 'faucet', 'shower', 'couch', 'food_jam', 'pillow', 'cup', 'novel', 'coffe_maker', 'coffee_filter', 'ice', 'curtain', 'kitchen_counter', 'spoon', 'clothes_shirt']\nTask:Give milk to cat\nresponse:\n", "output": {"thoughts": "Grab milk and pour it into the plate so that the cat can drink", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[POINTAT] <cat>", "[PULL] <cat>", "[FIND] <milk>", "[GRAB] <milk>", "[FIND] <plate>", "[POUR] <milk> <plate>", "[TOUCH] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'toy', 'television', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'table']\nTask:Pay bills\nresponse:\n{\n    \"thoughts\": Check my bills, on the computer, open the appropriate sites, pay my due bills,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['mop', 'walllamp', 'drawing', 'orchid', 'knife', 'floor', 'ceiling', 'folder', 'remote_control', 'food_cake', 'button', 'ceilinglamp', 'tvstand', 'mat', 'clothes_socks', 'pillow', 'dining_room', 'wall', 'powersocket', 'bed', 'toilet_paper', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'dishrack', 'clothes_dress', 'door', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the television. Switch on the television. Find the remote control. Grab the remote control. Point at television. Find the button. Push button.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'blow_dryer', 'walllamp', 'knife', 'orchid', 'cpuscreen', 'floor', 'computer', 'button', 'ceilinglamp', 'oil', 'mousepad', 'cup', 'bed', 'keys', 'novel', 'check', 'coffe_maker', 'toothbrush_holder', 'coffee_filter', 'food_food', 'curtain', 'pencil', 'shoes', 'controller', 'hairdryer', 'bathroom_cabinet', 'table', 'board_game']\nTask:Get ready to leave\nresponse:\n{\n    \"thoughts\": Get your jacket and shoes on, locate the keys off the side table, walk out the front door and secure building with keys when you leave,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>', '[FIND] <keys>', '[GRAB] <keys>']\n}\n", "input": "\nScene_Objects:['pot', 'measuring_cup', 'toilet', 'chair', 'band-aids', 'freezer', 'walllamp', 'food_carrot', 'toothbrush', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'button', 'mat', 'tvstand', 'faucet', 'mousepad', 'food_kiwi', 'wall', 'light_bulb', 'coffe_maker', 'toothbrush_holder', 'rag', 'pencil', 'sink', 'bathroom_cabinet', 'soap', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "getting ready for bed", "actions": ["[WALK] <bathroom_counter>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <toothbrush> <bathroom_counter>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[GRAB] <toothbrush>", "[SCRUB] <toothbrush>", "[RINSE] <toothbrush>", "[FIND] <toothbrush_holder>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'blender', 'chair', 'bathroom', 'hanger', 'floor', 'tea_bag', 'food_cake', 'food_bacon', 'piano_bench', 'mat', 'plate', 'dish_soap', 'mousepad', 'closetdrawer', 'couch', 'clothes_socks', 'wall', 'powersocket', 'coffe_maker', 'curtain', 'console', 'kitchen_counter', 'cupboard', 'door', 'sink', 'napkin', 'dishwasher', 'colander']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_bread', 'crayon', 'measuring_cup', 'chair', 'walllamp', 'trashcan', 'bedroom', 'bookshelf', 'purse', 'table_cloth', 'faucet', 'clothes_socks', 'pillow', 'wall', 'oven', 'newspaper', 'photoframe', 'coffe_maker', 'food_food', 'check', 'coffee_filter', 'shaving_cream', 'pencil', 'headset', 'dishrack', 'comb', 'dresser', 'spoon', 'instrument_guitar']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room, find lamp, switch on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'dough', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'tvstand', 'stereo', 'mousepad', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'sink', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to the bedroom and sit at my desk. I turn my radio on and listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'homework', 'window', 'wall', 'powersocket', 'oven', 'toy', 'bed', 'toilet_paper', 'television', 'toothbrush_holder', 'food_donut', 'tray', 'doorjamb', 'door', 'sink', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Lift the remote and change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'beer', 'microwave', 'chair', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'mop_bucket', 'toaster', 'brush', 'plate', 'faucet', 'stereo', 'ceilinglamp', 'tvstand', 'dish_soap', 'wall', 'food_food', 'video_game_controller', 'toothbrush_holder', 'doorjamb', 'mouse', 'kitchen_counter', 'headset', 'cupboard', 'sink', 'light', 'table']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": i go to the kitchen, i find the faucet. I turn the faucet on, i grab the plate to wash , i then get the soap. I wash the dishes by hand,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[WASH] <plate>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'walllamp', 'nail_polish', 'food_carrot', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'plate', 'mousepad', 'wall', 'powersocket', 'oven', 'food_food', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'napkin', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk Into Kitchen. Open Refrigerator. Grab Food. Walk to Microwave. Insert food into microwave. Cook until warm.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[CLOSE] <freezer>", "[WALK] <microwave>", "[SWITCHON] <microwave>", "[SWITCHOFF] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'plate', 'couch', 'pillow', 'wall', 'newspaper', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'bathtub', 'light', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": My cat looks like it wants attention, so I will go pet it on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'bookmark', 'stovefan', 'ceilinglamp', 'mat', 'mousepad', 'oil', 'food_peanut_butter', 'broom', 'wall', 'novel', 'diary', 'mouse', 'cupboard', 'clothes_dress', 'door', 'table', 'light', 'dirt']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go to office. Turn on computer. Pull out chair. Sit in chair. Use keyboard and mouse.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'juice', 'toilet', 'walllamp', 'toothbrush', 'drawing', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stamp', 'faucet', 'balanceball', 'clothes_pants', 'wall', 'powersocket', 'photoframe', 'novel', 'toothbrush_holder', 'television', 'food_donut', 'tray', 'towel_rack', 'cupboard', 'controller', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I go into the bathroom, go to the sink, get toothbrush and toothpaste. Put toothpaste on the toothbrush, brush teeth, turn on faucet, rinse brush, and return toothbrush.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['tea', 'cat', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'mail', 'alcohol', 'mat', 'faucet', 'clothes_pants', 'pillow', 'wall', 'wall_clock', 'oven', 'pasta', 'bench', 'coffee_filter', 'ice', 'paper_towel', 'conditioner', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk to the bathroom and find the sink. I grab the soap and rub some on my hands. I turn the faucet on and put my hands in the water. I rinse my hands and then turn off the faucet. Then I dry my hands on a paper towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <hands_both>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'orchid', 'floor', 'desk', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'pillow', 'nightstand', 'wall_clock', 'oven', 'wall', 'coffe_maker', 'mouthwash', 'television', 'tray', 'towel_rack', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to phone, grab phone, pick up phone, put to ear,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['juice', 'food_salt', 'cd_player', 'toilet', 'freezer', 'walllamp', 'nail_polish', 'drawing', 'orchid', 'knifeblock', 'floor', 'phone', 'keyboard', 'computer', 'button', 'toaster', 'ceilinglamp', 'window', 'couch', 'pillow', 'bowl', 'wall', 'oven', 'photoframe', 'check', 'curtain', 'kettle', 'pajamas', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room, walk to cd player, turn on cd player, press play button, walk to sofa, sit in sofa and listen to music", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <button>", "[PUSH] <button>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'wooden_spoon', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to my home office. I sit on the chair right in front of my computer. I turn on the computer. Use the mouse to open the web browser. Use the keyboard to type the URL that I want to visit.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['toilet', 'bathroom', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'alcohol', 'mat', 'electric_shaver', 'oil', 'wall', 'powersocket', 'box', 'food_cheese', 'photoframe', 'bed', 'food_food', 'coffe_maker', 'doorjamb', 'lighter', 'dishrack', 'door', 'bathroom_cabinet', 'light', 'dishwasher', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "push any ligth button to turn it on, for example in the living room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'character', 'shoe-shine_kit', 'faucet', 'couch', 'form', 'wall', 'bench', 'toy', 'cup', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom, find sink, turn on faucet, put soap in hands, wash and rinse hands,,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <soap>']\n}\n", "input": "\nScene_Objects:['juice', 'microwave', 'toilet', 'freezer', 'bathroom', 'orchid', 'knifeblock', 'phone', 'character', 'stovefan', 'mat', 'shower', 'dustpan', 'food_orange', 'pillow', 'broom', 'box', 'newspaper', 'check', 'television', 'coffee_pot', 'controller', 'door', 'home_office', 'table', 'bathroom_cabinet', 'dishwasher', 'colander', 'teeth', 'shoe_rack']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter the Home office, then walk to the phone, pickup the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'nail_polish', 'bathroom', 'cat', 'orchid', 'cpuscreen', 'longboard', 'bedroom', 'keyboard', 'clothes_jacket', 'bookmark', 'toaster', 'balanceball', 'stereo', 'mousepad', 'clothes_pants', 'pillow', 'nightstand', 'wall', 'photoframe', 'food_food', 'mouthwash', 'tray', 'diary', 'comb', 'pajamas', 'maindoor', 'bag', 'spoon', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Go to the living room where the cat sleeps. Find cat. Proceed to petting cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['freezer', 'band-aids', 'dough', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'dvd_player', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'powersocket', 'cup', 'bed', 'photoframe', 'newspaper', 'water', 'food_donut', 'doorjamb', 'mouse', 'door', 'light', 'table', 'shoe_rack']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <freezer>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <cup>", "[POUR] <water> <cup>", "[FIND] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'walllamp', 'cat', 'ceilingfan', 'drawing', 'cpuscreen', 'clothes_underwear', 'longboard', 'floor', 'ceiling', 'keyboard', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'oil', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'ice', 'curtain', 'after_shave', 'door', 'table', 'dishwasher', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on couch and watch the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'mop', 'bathroom', 'cat', 'trashcan', 'knifeblock', 'stamp', 'button', 'piano_bench', 'faucet', 'mousepad', 'bowl', 'nightstand', 'wall_clock', 'oven', 'wall', 'bench', 'photoframe', 'novel', 'food_food', 'coffe_maker', 'curtain', 'pencil', 'shoes', 'hairbrush', 'hair', 'instrument_guitar', 'dishwasher', 'tooth_paste']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up  the book open the book and read", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'band-aids', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'homework', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'keys', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'light', 'soap']\nTask:Change TV channels\nresponse:\n{\n    \"thoughts\": Take remote. Change to desired channel by you.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'chair', 'freezer', 'drawing', 'knifeblock', 'cpuscreen', 'needle', 'floor', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'brush', 'mousepad', 'wall', 'nightstand', 'bed', 'photoframe', 'check', 'food_food', 'doorjamb', 'mouse', 'kitchen_counter', 'headset', 'door', 'table', 'light', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I'll walk to my bedroom, where I have my computer. I'll walk to my computer and sit down in front of it. I'll put on my headphones and unlock my computer. I'll then navigate to music stored locally or streamed through the web.", "actions": ["[WALK] <bedroom>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[WALK] <computer>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <headset>", "[GRAB] <headset>", "[PUTON] <headset>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'jelly', 'drawing', 'orchid', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'computer', 'bookmark', 'faucet', 'clothes_socks', 'couch', 'pillow', 'broom', 'wall', 'wall_clock', 'tape', 'photoframe', 'novel', 'check', 'food_onion', 'doorjamb', 'door', 'bathtub', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room and relax on the sofa with a book,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[DROP] <bookmark>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['cd_player', 'walllamp', 'drawing', 'knife', 'knifeblock', 'ceiling', 'character', 'keyboard', 'ceilinglamp', 'tvstand', 'stereo', 'shower', 'faucet', 'window', 'food_noodles', 'mousepad', 'couch', 'dustpan', 'dining_room', 'food_kiwi', 'wall_clock', 'bench', 'powersocket', 'oven', 'toy', 'photoframe', 'ice', 'dishrack', 'clothes_skirt', 'coffee']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk in the living room, then I go the the stereo and the cd player and turn them on.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'crayon', 'chair', 'cat', 'dough', 'milk', 'hanger', 'floor', 'keyboard', 'mail', 'table_cloth', 'stovefan', 'plate', 'faucet', 'tvstand', 'window', 'couch', 'form', 'dining_room', 'cutting_board', 'photoframe', 'check', 'wine', 'dresser', 'hairbrush', 'pajamas', 'conditioner', 'home_office', 'bathroom_cabinet', 'table']\nTask:Give milk to cat\nresponse:\n{\n    \"thoughts\": Grab milk and pour it into the plate so that the cat can drink,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[POINTAT] <cat>', '[PULL] <cat>', '[FIND] <milk>', '[GRAB] <milk>', '[FIND] <plate>', '[POUR] <milk> <plate>', '[TOUCH] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'band-aids', 'trashcan', 'dough', 'food_egg', 'cpuscreen', 'floor', 'ceiling', 'folder', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'pillow', 'wall', 'powersocket', 'newspaper', 'food_food', 'check', 'oven_mitts', 'food_donut', 'food_onion', 'doorjamb', 'kettle', 'mouse', 'pajamas', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "go to the kitchen and put the groceries in the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'microwave', 'walllamp', 'toothbrush', 'drawing', 'knifeblock', 'wallshelf', 'keyboard', 'food_cake', 'ceilinglamp', 'tvstand', 'shower', 'stereo', 'oil', 'tooth_paste', 'couch', 'closetdrawer', 'dining_room', 'toy', 'basket_for_clothes', 'light_bulb', 'check', 'centerpiece', 'kettle', 'mouse', 'kitchen_counter', 'shoes', 'light', 'board_game']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": EARLY MORNING AND BEFORE GOING TO BED,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'scissors', 'mop', 'microwave', 'toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'tvstand', 'shower', 'stereo', 'mousepad', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'chessboard', 'laundry_detergent', 'door', 'board_game']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "walk to bathroom , run to toilet", "actions": ["[WALK] <bathroom>", "[RUN] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'microwave', 'freezer', 'walllamp', 'blow_dryer', 'detergent', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'faucet', 'mousepad', 'wall', 'wall_clock', 'oven', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'cards', 'kitchen_counter', 'cupboard', 'sink', 'light', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, grab a cup, fill with water and then drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[FIND] <sink>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'beer', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'foundation', 'window', 'wall', 'powersocket', 'cup', 'photoframe', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk into the Living room.  Find the television remote.  Use it to turn on the television.  Point the remote at the television. Press the channel button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'microwave', 'chair', 'freezer', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'stereo', 'picture', 'mat', 'closetdrawer', 'couch', 'dining_room', 'wall_clock', 'bench', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'home_office', 'instrument_guitar']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the living room and then I walk to the light switch in there and turn the lights on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'chair', 'walllamp', 'food_butter', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'shower', 'picture', 'mousepad', 'pillow', 'wall', 'oven', 'powersocket', 'bed', 'photoframe', 'food_donut', 'tray', 'mouse', 'pencil', 'bag', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to living room,find the text book on the self ,sat on the chair ,open and read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'floor', 'alcohol', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'deck_of_cards', 'window', 'shower', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'photoframe', 'keys', 'television', 'curtain', 'doorjamb', 'mouse', 'home_office', 'table', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into bedrrom and to the sofa. Sit on the sofa and touch cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'trashcan', 'drawing', 'floor', 'phone', 'ceiling', 'desk', 'purse', 'mat', 'tvstand', 'ceilinglamp', 'window', 'food_peanut_butter', 'food_jam', 'pillow', 'wall', 'bench', 'check', 'toothbrush_holder', 'colander', 'tray', 'rag', 'lighter', 'kitchen_counter', 'comb', 'door', 'table', 'dishwasher', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Pick up phone", "actions": ["[WALK] <table>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'crayon', 'ceilingfan', 'drawing', 'needle', 'floor', 'ceiling', 'shampoo', 'keyboard', 'remote_control', 'thread', 'food_bacon', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'photoframe', 'bed', 'television', 'coffee_filter', 'curtain', 'mouse', 'table', 'light', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go into the living room. I find the remote. I turn on the tv. I sit on the couch and watch tv.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'freezer', 'knifeblock', 'knife', 'bathroom_counter', 'shampoo', 'ceiling', 'keyboard', 'stovefan', 'mat', 'stereo', 'clothes_socks', 'oil', 'nightstand', 'powersocket', 'toy', 'novel', 'food_food', 'coffe_maker', 'rag', 'kettle', 'shoes', 'cupboard', 'hair', 'controller', 'door', 'instrument_guitar', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk into living room, walk to stereo, turn on stereo, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_pants', 'form', 'pillow', 'towel', 'wall', 'powersocket', 'bench', 'newspaper', 'television', 'curtain', 'doorjamb', 'mouse', 'hair', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'food_pizza', 'walllamp', 'cat', 'toothbrush', 'orchid', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'food_bacon', 'table_cloth', 'plate', 'food_jam', 'bench', 'oven', 'tape', 'photoframe', 'toothbrush_holder', 'oven_mitts', 'wooden_spoon', 'diary', 'mouse', 'sink', 'home_office', 'light', 'tooth_paste', 'clothes_shirt']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to over to desk, sit down on chair and face the computer, turn the computer on, use arms to place hands on the keyboard, type work into the computer and use eyes to review.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'tablelamp', 'freezer', 'bathroom', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'mousepad', 'food_jam', 'pillow', 'wall', 'powersocket', 'light_bulb', 'drinking_glass', 'food_onion', 'doorjamb', 'mouse', 'cd', 'door', 'light', 'clothes_skirt', 'coffee']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Robot Drinks the water", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>", "[PUTOBJBACK] <drinking_glass>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'mouse', 'dresser', 'conditioner', 'door', 'table', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": For relaxing walk to living room. Find the sofa to relax. Arrange the sofa to relax. Sit on the sofa to relaxing.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <conditioner>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'band-aids', 'bathroom', 'trashcan', 'phone', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'picture', 'window', 'couch', 'bowl', 'broom', 'wall', 'oven', 'bench', 'cup', 'food_food', 'oven_mitts', 'colander', 'curtain', 'pencil', 'kitchen_counter', 'clothes_dress', 'sink', 'dishwasher', 'soap', 'instrument_violin']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find Cat in living room.  Walk to cat.  Pet cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'folder', 'character', 'desk', 'faucet', 'stereo', 'mousepad', 'food_orange', 'wall', 'oven', 'photoframe', 'oven_mitts', 'water', 'drinking_glass', 'wooden_spoon', 'towel_rack', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'spoon', 'board_game']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Go to kitchen, open cabinet, get glass, close cabinet, put glass under faucet, turn on water, turn off water when glass has 85% water, lift glass to mouth, open mouth and put on glass, tilt glass up and swallow water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[TURNTO] <kitchen_cabinet>', '[LOOKAT] <kitchen_cabinet>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[FIND] <sink>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[SWITCHOFF] <faucet>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['glue', 'tea', 'microwave', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'phone', 'toaster', 'shower', 'faucet', 'food_noodles', 'mousepad', 'microphone', 'pillow', 'wall', 'bed', 'light_bulb', 'coffe_maker', 'toothbrush_holder', 'coffee_filter', 'oven_mitts', 'razor', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'dustpan', 'pillow', 'wall', 'wall_clock', 'powersocket', 'television', 'tray', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Browsing the internet for information.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'cd_player', 'walllamp', 'trashcan', 'food_carrot', 'drawing', 'needle', 'bathroom_counter', 'dvd_player', 'bedroom', 'bookshelf', 'alcohol', 'piano_bench', 'bookmark', 'laptop', 'stereo', 'couch', 'pillow', 'broom', 'bench', 'basket_for_clothes', 'mouse', 'shoes', 'headset', 'food_steak', 'sink', 'hairdryer', 'instrument_guitar', 'door', 'bag']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the playback device you want to listen to. turn on. press play. enjoyu music.", "actions": ["[WALK] <bedroom>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'food_salt', 'chair', 'walllamp', 'detergent', 'drawing', 'clothes_underwear', 'floor', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'food_noodles', 'garbage_can', 'mousepad', 'microphone', 'wall', 'nightstand', 'food_vegetable', 'bed', 'food_food', 'coffee_filter', 'mouse', 'sink', 'home_office', 'table', 'dishwasher', 'pantry']\nTask:Organize pantry\nresponse:\n{\n    \"thoughts\": Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <pantry>', '[FIND] <food_food>', '[GRAB] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[FIND] <garbage_can>', '[PUTBACK] <food_food> <garbage_can>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <dry_pasta>', '[GRAB] <dry_pasta>', '[TURNTO] <dry_pasta>', '[LOOKAT] <dry_pasta>', '[PUTOBJBACK] <dry_pasta>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[PUSH] <food_food>', '[FIND] <food_noodles>', '[GRAB] <food_noodles>', '[TURNTO] <food_noodles>', '[LOOKAT] <food_noodles>', '[PUTOBJBACK] <food_noodles>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'walllamp', 'trashcan', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'tea_bag', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'deck_of_cards', 'homework', 'wall', 'powersocket', 'check', 'television', 'doorjamb', 'mouse', 'cupboard', 'controller', 'door', 'bag', 'table', 'teeth']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Go to the living room, Switch on the television. Find the remote and press the change channel button, change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[TOUCH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'food_salt', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'computer', 'mat', 'faucet', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'toothbrush_holder', 'wine', 'doorjamb', 'shaving_cream', 'kitchen_counter', 'door', 'sink', 'clothes_scarf', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into living room. Find wall. Find light button. Switch on light button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'plate', 'stereo', 'mousepad', 'wall', 'bench', 'television', 'wine', 'rag', 'lighter', 'shaving_cream', 'mouse', 'hairbrush', 'door', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "walk to living room, look at computer, find chair, sit in chair, grab keyboard, grab mouse, type using keyboard and mouse", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[WALK] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'jelly', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'homework', 'closetdrawer', 'couch', 'pillow', 'wall', 'newspaper', 'coffe_maker', 'television', 'tray', 'curtain', 'towel_rack', 'cupboard', 'home_office', 'table']\nTask:Read newspaper\nresponse:\n{\n    \"thoughts\": Read the newspaper on the couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <newspaper>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'remote_control', 'mop_bucket', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'cup', 'newspaper', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'shoes', 'dresser', 'door', 'home_office', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'bathroom', 'hanger', 'bathroom_counter', 'man', 'bookshelf', 'keyboard', 'piano_bench', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'window', 'closetdrawer', 'dining_room', 'photoframe', 'tray', 'curtain', 'food_onion', 'mouse', 'dresser', 'cupboard', 'door', 'sink', 'home_office', 'light', 'clothes_skirt', 'bathroom_cabinet']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": When someone knocks I n the door walk through the hallway and greet them.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'orchid', 'kitchen_cabinet', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'toaster', 'stovefan', 'faucet', 'mousepad', 'couch', 'wall', 'pasta', 'oven', 'box', 'food_food', 'water', 'tray', 'pencil', 'kitchen_counter', 'sauce_pan', 'cupboard', 'instrument_guitar', 'bathroom_cabinet']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Cook some pasta on the stove", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pasta>", "[GRAB] <pasta>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <sauce_pan>", "[POUR] <water> <sauce_pan>", "[FIND] <oven>", "[SWITCHON] <oven>", "[PUTBACK] <pasta> <sauce_pan>", "[FIND] <sauce_pan>", "[GRAB] <sauce_pan>", "[PUTBACK] <sauce_pan> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'microwave', 'toilet', 'freezer', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bookshelf', 'computer', 'toaster', 'mat', 'shower', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'bench', 'cup', 'food_food', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'napkin']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Find groceries. Grab groceries. Walk to refrigerator. Open refrigerator door. Place groceries into refrigerator. Close refrigerator door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tea', 'toilet', 'dog', 'trashcan', 'orchid', 'drawing', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'alcohol', 'toaster', 'mat', 'shower', 'picture', 'stereo', 'clothes_socks', 'wall', 'nightstand', 'photoframe', 'check', 'razor', 'curtain', 'doorjamb', 'pencil', 'door', 'napkin', 'laser_pointer']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "i go to the bathroom. I find the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'cat', 'drawing', 'floor', 'dvd_player', 'bookshelf', 'food_cake', 'keyboard', 'computer', 'bookmark', 'mat', 'shower', 'faucet', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'television', 'curtain', 'towel_rack', 'mouse', 'door', 'sink', 'bathtub', 'light']\nTask:Breakfast\nresponse:\n{\n    \"thoughts\": I have breakfast in bed, watching tv.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[GRAB] <food_food>', '[WALK] <bedroom>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <bed>', '[LIE] <bed>', '[TURNTO] <television>', '[WATCH] <television>', '[EAT] <food_food>']\n}\n", "input": "\nScene_Objects:['water_glass', 'mop', 'freezer', 'bathroom', 'trashcan', 'knifeblock', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'table_cloth', 'stovefan', 'mat', 'mousepad', 'pillow', 'bench', 'oven', 'photoframe', 'bed', 'coffee_filter', 'wine', 'centerpiece', 'dresser', 'sink', 'table', 'tooth_paste', 'laser_pointer', 'teeth']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, walk to table, grab water glass and drink water", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>", "[PUTBACK] <water_glass> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'jelly', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'food_chicken', 'desk', 'keyboard', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'stovefan', 'faucet', 'mousepad', 'oil', 'wall', 'powersocket', 'oven', 'food_food', 'tray', 'doorjamb', 'coffee_pot', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'walllamp', 'food_carrot', 'milk', 'drawing', 'knifeblock', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'mat', 'stereo', 'stovefan', 'faucet', 'mousepad', 'oil', 'pillow', 'wall', 'powersocket', 'bench', 'oven', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'napkin', 'light', 'table']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and open the fridge. From the fridge take some chicken and close the fridge. Walk to the table and put the chicken on the table. Plug in oven and put the chicken into the oven.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <food_chicken> <table>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <table>", "[FIND] <oven>", "[PLUGIN] <oven>", "[OPEN] <oven>", "[GRAB] <food_chicken>", "[PUTIN] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'food_ice_cream', 'mop', 'chair', 'freezer', 'walllamp', 'cat', 'trashcan', 'toothbrush', 'hanger', 'bathroom_counter', 'floor', 'keyboard', 'dining_room', 'bench', 'powersocket', 'cup', 'photoframe', 'novel', 'toothbrush_holder', 'curtain', 'mouse', 'headset', 'pajamas', 'door', 'dishwasher', 'vase', 'tooth_paste', 'clothes_shirt', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the toothbrush holder, grab the toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now wash the teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'microwave', 'toilet', 'drawing', 'orchid', 'floor', 'phone', 'tea_bag', 'bookshelf', 'plate', 'shower', 'faucet', 'oil', 'candle', 'form', 'dining_room', 'wall', 'basket_for_clothes', 'photoframe', 'bed', 'food_food', 'coffee_filter', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'door', 'dishwasher']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bath room and go to the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'bathroom', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'knifeblock', 'floor', 'dvd_player', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'faucet', 'electric_shaver', 'floor_lamp', 'mousepad', 'wall', 'powersocket', 'novel', 'food_food', 'check', 'television', 'after_shave', 'cupboard', 'sink', 'home_office', 'table', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Twist the lamp switch until the light goes on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'food_bread', 'microwave', 'walllamp', 'trashcan', 'knife', 'cpuscreen', 'floor', 'phone', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'plate', 'shower', 'foundation', 'clothes_pants', 'wall', 'coffe_maker', 'television', 'water', 'kitchen_counter', 'hairbrush', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "first i will take all the dishes to the dish washer and with the help of dishwasher and water i will clean all the dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water>", "[TURNTO] <water>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'toilet', 'dough', 'toothbrush', 'drawing', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'stereo', 'shower', 'window', 'oil', 'couch', 'bowl', 'wall', 'oven', 'toy', 'cup', 'bed', 'food_food', 'television', 'wooden_spoon', 'towel_rack', 'comb', 'cupboard', 'table', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Find couch. Sit on couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'tablelamp', 'mop', 'toilet', 'walllamp', 'cat', 'drawing', 'floor', 'ceiling', 'character', 'bookshelf', 'food_cake', 'piano_bench', 'shower', 'faucet', 'pillow', 'wall', 'photoframe', 'bed', 'curtain', 'doorjamb', 'diary', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'soap', 'instrument_violin']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[OPEN] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'microwave', 'toilet', 'chair', 'freezer', 'band-aids', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'food_bacon', 'mat', 'tvstand', 'shower', 'pillow', 'wall', 'bed', 'photoframe', 'food_food', 'doorjamb', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'door']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'crayon', 'chair', 'freezer', 'detergent', 'food_butter', 'drawing', 'knife', 'hanger', 'floor', 'character', 'piano_bench', 'toaster', 'clothes_socks', 'food_peanut_butter', 'pillow', 'food_kiwi', 'coffe_maker', 'food_donut', 'tray', 'curtain', 'shaving_cream', 'mouse', 'kitchen_counter', 'dishrack', 'hair', 'door', 'hairdryer', 'napkin']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'microwave', 'blow_dryer', 'detergent', 'cat', 'drawing', 'desk', 'thread', 'computer', 'button', 'table_cloth', 'ceilinglamp', 'tvstand', 'mousepad', 'broom', 'nightstand', 'oven', 'powersocket', 'box', 'newspaper', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'conditioner', 'bag', 'napkin', 'table', 'light', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my cat, and I pet him,\n    \"actions\": ['[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'food_ice_cream', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'mousepad', 'food_peanut_butter', 'food_orange', 'wall', 'oven', 'powersocket', 'cutting_board', 'water', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'spoon', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I enter the kitchen and grab a waterglass from the cupboard. I open the fridge and grap some water. I pour the water into the cup and drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[OPEN] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'chair', 'detergent', 'milk', 'knife', 'drawing', 'cpuscreen', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'faucet', 'couch', 'wall', 'photoframe', 'bed', 'water', 'after_shave', 'mouse', 'controller', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Walk to sink and turn on the water. Grab the soap and wash the hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <water>', '[TURNTO] <water>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_salt', 'chair', 'jelly', 'walllamp', 'milk', 'floor', 'bedroom', 'keyboard', 'button', 'plate', 'tvstand', 'shower', 'window', 'dustpan', 'dining_room', 'nightstand', 'oven', 'basket_for_clothes', 'check', 'coffe_maker', 'tray', 'controller', 'clothes_dress', 'food_steak', 'napkin', 'soap', 'clothes_shirt', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'food_butter', 'drawing', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'tvstand', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'photoframe', 'bed', 'coffe_maker', 'television', 'food_donut', 'tray', 'curtain', 'doorjamb', 'pencil', 'dresser', 'napkin', 'dirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'bathroom', 'orchid', 'knifeblock', 'cpuscreen', 'needle', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'mousepad', 'couch', 'dining_room', 'wall', 'basket_for_clothes', 'light_bulb', 'food_food', 'check', 'toothbrush_holder', 'mouse', 'comb', 'door', 'clothes_scarf', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk to the home office. I open the door of the office room and sit on the chair and in front of me is my desktop computer,  and I switch on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'orchid', 'drawing', 'ceiling', 'bedroom', 'computer', 'mat', 'shoe-shine_kit', 'stereo', 'ceilinglamp', 'window', 'tvstand', 'food_orange', 'nightstand', 'bench', 'bed', 'coffe_maker', 'food_food', 'television', 'doorjamb', 'mouse', 'pencil', 'dresser', 'pajamas', 'cupboard', 'bathtub', 'table', 'clothes_shirt', 'coffee']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'food_bacon', 'toaster', 'tvstand', 'window', 'oil', 'couch', 'wall', 'nightstand', 'newspaper', 'coffe_maker', 'television', 'coffee_filter', 'water', 'food_donut', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'dirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'bookmark', 'toaster', 'mat', 'tvstand', 'faucet', 'shower', 'laptop', 'mousepad', 'wall', 'oven', 'powersocket', 'towel_rack', 'kitchen_counter', 'headset', 'door', 'sink', 'bathtub', 'home_office', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I go to the bathroom. I locate the toilet. I walk to the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[WALK] <toilet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'toilet', 'chair', 'trashcan', 'drawing', 'food_egg', 'floor', 'ceiling', 'bookshelf', 'mat', 'ceilinglamp', 'shower', 'food_jam', 'dining_room', 'broom', 'wall', 'bench', 'photoframe', 'curtain', 'food_onion', 'doorjamb', 'dresser', 'hairbrush', 'cupboard', 'door', 'food_steak', 'light', 'table', 'teeth']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to the bathroom and wash with the shower.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[WASH] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'freezer', 'ground_coffee', 'walllamp', 'food_egg', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'stovefan', 'plate', 'mousepad', 'wall', 'nightstand', 'powersocket', 'novel', 'coffe_maker', 'check', 'coffee_filter', 'water', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'soap', 'teeth', 'shoe_rack']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'scissors', 'microwave', 'jelly', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'shampoo', 'ceiling', 'keyboard', 'mail', 'brush', 'toaster', 'stovefan', 'plate', 'picture', 'oil', 'form', 'bowl', 'photoframe', 'newspaper', 'bed', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'home_office']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to living room. Find book. Grab book. Open book. Read book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'tea_bag', 'bookshelf', 'food_bacon', 'toaster', 'plate', 'tvstand', 'dish_soap', 'mousepad', 'couch', 'wall', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'food_food', 'toothbrush_holder', 'ice', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'clothes_dress', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['glue', 'mop', 'microwave', 'trashcan', 'bills', 'orchid', 'knifeblock', 'drawing', 'hanger', 'floor', 'phone', 'tvstand', 'form', 'pillow', 'bench', 'powersocket', 'light_bulb', 'video_game_controller', 'water', 'centerpiece', 'doorjamb', 'mouse', 'pencil', 'cd', 'door', 'napkin', 'dishwasher', 'soap', 'hands_both', 'creditcard']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the water, wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <water>", "[FIND] <water>", "[TURNTO] <water>", "[LOOKAT] <water>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'glue', 'filing_cabinet', 'microwave', 'jelly', 'freezer', 'ceilingfan', 'orchid', 'food_egg', 'floor', 'phone', 'ceiling', 'desk', 'thread', 'mat', 'laptop', 'deck_of_cards', 'faucet', 'foundation', 'pillow', 'nightstand', 'powersocket', 'basket_for_clothes', 'food_food', 'lighter', 'sink', 'home_office', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'bathroom', 'hanger', 'floor', 'bookshelf', 'table_cloth', 'stereo', 'window', 'electric_shaver', 'mousepad', 'couch', 'food_orange', 'oven', 'photoframe', 'newspaper', 'bed', 'doorjamb', 'kettle', 'cards', 'hairbrush', 'dresser', 'food_steak', 'clothes_scarf', 'napkin', 'home_office', 'table', 'clothes_skirt', 'scrabble', 'teeth']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'ground_coffee', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'toaster', 'plate', 'faucet', 'foundation', 'mousepad', 'food_jam', 'food_kiwi', 'wall', 'pasta', 'powersocket', 'food_food', 'coffe_maker', 'rag', 'shaving_cream', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Take a dish and put it in the dishwasher to be washed,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'microwave', 'walllamp', 'food_carrot', 'character', 'desk', 'keyboard', 'shower', 'washing_machine', 'window', 'mousepad', 'clothes_pants', 'towel', 'wall', 'powersocket', 'basket_for_clothes', 'check', 'mouthwash', 'food_rice', 'headset', 'chessboard', 'hair', 'napkin', 'table', 'soap', 'scrabble', 'dirt', 'coffee', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[WALK] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'trashcan', 'toothbrush', 'bathroom_counter', 'shampoo', 'floor', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'toaster', 'ceilinglamp', 'iron', 'shower', 'stereo', 'pillow', 'wall', 'wall_clock', 'photoframe', 'food_food', 'water', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[FIND] <toothbrush>', '[FIND] <water>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'microwave', 'ground_coffee', 'walllamp', 'detergent', 'trashcan', 'food_butter', 'drawing', 'knifeblock', 'floor', 'ceiling', 'toaster', 'mat', 'food_jam', 'pillow', 'wall', 'oven', 'bench', 'cup', 'newspaper', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk into the kitchen. I find the coffee maker and open it. I find the coffee filter, grab it and put it into the coffee maker. I find ground coffee, grab it and put it into the coffee maker. I find some water, then I pour it into the coffee maker. I close the coffee maker, I plug it in and I switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I want to surf the internet and look around at funny pictures.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'microwave', 'chair', 'bathroom', 'orchid', 'drawing', 'food_egg', 'knife', 'floor', 'computer', 'tvstand', 'window', 'couch', 'candle', 'microphone', 'pillow', 'towel', 'wall', 'food_food', 'check', 'television', 'wooden_spoon', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'napkin', 'table', 'coffee']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "walk to entrance hall. find a table. pull table. walk to living rook. push table.", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[PULL] <table>", "[WALK] <home_office>", "[WALK] <table>", "[PUSH] <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'crayon', 'walllamp', 'trashcan', 'dough', 'floor', 'phone', 'bedroom', 'tea_bag', 'toaster', 'mat', 'plate', 'ceilinglamp', 'dish_soap', 'wall', 'pasta', 'oven', 'cutting_board', 'food_cheese', 'newspaper', 'coffe_maker', 'check', 'food_food', 'tray', 'towel_rack', 'pencil', 'dresser', 'conditioner', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'milk', 'floor', 'ceiling', 'desk', 'keyboard', 'button', 'mop_bucket', 'toaster', 'stereo', 'window', 'wall', 'powersocket', 'bench', 'newspaper', 'coffe_maker', 'tray', 'curtain', 'food_sugar', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'coffee_cup', 'dishwasher', 'light', 'clothes_shirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[LOOKAT] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <milk>", "[GRAB] <milk>", "[PUTBACK] <milk> <coffe_maker>", "[FIND] <food_sugar>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[SWITCHON] <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'detergent', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'shoe-shine_kit', 'faucet', 'clothes_socks', 'closetdrawer', 'mousepad', 'dining_room', 'broom', 'wall', 'nightstand', 'photoframe', 'television', 'water', 'tray', 'cards', 'paper_towel', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. I turn on the water. I put soap on my hands. I rinse my hands. I turn off the water. I dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <water>', '[TURNTO] <water>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <water>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'mop', 'tea', 'freezer', 'walllamp', 'floor', 'character', 'stovefan', 'toaster', 'mat', 'tvstand', 'stereo', 'window', 'electric_shaver', 'food_noodles', 'couch', 'soap', 'pillow', 'wall', 'bench', 'toilet_paper', 'food_food', 'doorjamb', 'diary', 'dresser', 'hairbrush', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to bathroom. find closet. Uset It. Clean it. Use Toilet paper", "actions": ["[WALK] <bathroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[CLOSE] <dresser>", "[FIND] <toilet_paper>", "[WASH] <dresser>", "[FIND] <toilet_paper>", "[TOUCH] <toilet_paper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'bathroom', 'food_carrot', 'wallshelf', 'ceiling', 'dvd_player', 'character', 'bedroom', 'desk', 'keyboard', 'thread', 'stovefan', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'cup', 'bed', 'food_food', 'razor', 'toothbrush_holder', 'curtain', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'napkin', 'table', 'clothes_skirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Go to the living room.  Find the telephone.  Lift up the receiver.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sponge', 'spectacles', 'microwave', 'walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'character', 'desk', 'alcohol', 'stovefan', 'ceilinglamp', 'mat', 'window', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'cutting_board', 'photoframe', 'light_bulb', 'television', 'curtain', 'kettle', 'napkin', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'keyboard', 'purse', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'food_noodles', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": FIRST YOU NEED TO SWITVH ON YOU COMPUTER AND CHECK THE INTERNET CONNECTION.FOR SENDING EMAIL EMAIL ID FOR REQUIRED FOR BOTH SEND AND RECEIVE .SO CLEALY MENTION THE ID ON THE TO PLACE THEN ONLY IT WILL GO CORRECTLY,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'detergent', 'trashcan', 'floor', 'ceiling', 'bedroom', 'toaster', 'mat', 'stereo', 'pillow', 'wall', 'cup', 'bed', 'coffe_maker', 'food_food', 'coffee_filter', 'water', 'food_donut', 'wooden_spoon', 'food_onion', 'doorjamb', 'pencil', 'kitchen_counter', 'cupboard', 'napkin', 'spoon', 'dishwasher', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I find the coffee machine, fill it with water and coffee then press the button to make the coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'juice', 'nail_polish', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'mat', 'shower', 'window', 'closetdrawer', 'wall', 'nightstand', 'cutting_board', 'bed', 'photoframe', 'coffe_maker', 'television', 'curtain', 'mouse', 'pencil', 'vacuum_cleaner', 'hair', 'maindoor', 'table', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I lie on my bed and read a book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <bed>', '[LIE] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'drawing', 'orchid', 'floor', 'folder', 'mop_bucket', 'ceilinglamp', 'stereo', 'picture', 'window', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'bench', 'powersocket', 'novel', 'coffe_maker', 'toilet_paper', 'toothbrush_holder', 'tray', 'curtain', 'lighter', 'headset', 'cupboard', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and find the sofa. I sit on the sofa and get comfortable.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'walllamp', 'drawing', 'hanger', 'floor', 'mat', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'cutting_board', 'cup', 'novel', 'food_food', 'television', 'wine', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'shoes', 'bathtub', 'bathroom_cabinet', 'table', 'soap', 'board_game']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": lie in the sofa, relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'walllamp', 'ground_coffee', 'trashcan', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'foundation', 'couch', 'microphone', 'wall', 'powersocket', 'oven', 'bed', 'television', 'wooden_spoon', 'doorjamb', 'lighter', 'dresser', 'door', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'microwave', 'chair', 'freezer', 'walllamp', 'detergent', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'stereo', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door', 'bag', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['scissors', 'food_ice_cream', 'chair', 'bathroom', 'trashcan', 'food_carrot', 'shampoo', 'phone', 'floor', 'stamp', 'bookshelf', 'ceilinglamp', 'window', 'homework', 'couch', 'clothes_pants', 'pillow', 'nightstand', 'cup', 'bed', 'check', 'food_food', 'ice', 'headset', 'cupboard', 'door', 'home_office', 'light', 'dishwasher', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to phone, grab phone, pick up phone, put to ear", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'filing_cabinet', 'food_salt', 'microwave', 'blow_dryer', 'detergent', 'knife', 'orchid', 'bathroom_counter', 'shampoo', 'keyboard', 'bookshelf', 'mail', 'stovefan', 'ceilinglamp', 'toaster', 'mat', 'faucet', 'food_peanut_butter', 'dining_room', 'nightstand', 'oven', 'cup', 'light_bulb', 'food_food', 'television', 'food_onion', 'pencil', 'napkin', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'scissors', 'toilet', 'freezer', 'ground_coffee', 'detergent', 'food_egg', 'cpuscreen', 'phone', 'folder', 'character', 'keyboard', 'envelope', 'piano_bench', 'ceilinglamp', 'plate', 'faucet', 'couch', 'coffe_maker', 'food_donut', 'tray', 'centerpiece', 'doorjamb', 'mouse', 'dishrack', 'dresser', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office, find where the phone is and lift the reciever.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'dough', 'orchid', 'cpuscreen', 'hanger', 'floor', 'keyboard', 'remote_control', 'food_cake', 'envelope', 'computer', 'button', 'food_bacon', 'ceilinglamp', 'stereo', 'homework', 'oven', 'tape', 'photoframe', 'coffee_table', 'keys', 'check', 'novel', 'light_bulb', 'wooden_spoon', 'curtain', 'cd', 'home_office']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_bread', 'microwave', 'freezer', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'box', 'oven_mitts', 'food_onion', 'doorjamb', 'kitchen_counter', 'chessboard', 'dresser', 'cupboard', 'door', 'light', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <bathroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_pizza', 'food_salt', 'jelly', 'band-aids', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'broom', 'wall', 'powersocket', 'doorjamb', 'lighter', 'door', 'sink', 'light', 'table', 'laser_pointer', 'clothes_shirt', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go into my room, find the light,  then turn on the lights.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'foundation', 'dustpan', 'pillow', 'broom', 'wall', 'powersocket', 'cup', 'video_game_controller', 'television', 'doorjamb', 'hairbrush', 'door', 'sink', 'bag', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'food_pizza', 'filing_cabinet', 'toilet', 'walllamp', 'detergent', 'orchid', 'food_egg', 'floor', 'phone', 'ceiling', 'character', 'keyboard', 'alcohol', 'toaster', 'tvstand', 'faucet', 'balanceball', 'picture', 'wall', 'oven', 'tape', 'food_cheese', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'towel_rack', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the sink. Take dishes out of sink. Open dishwasher and place dishes in dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tea', 'walllamp', 'drawing', 'orchid', 'food_egg', 'floor', 'ceiling', 'phone', 'character', 'remote_control', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'shower', 'couch', 'clothes_hat', 'wall', 'powersocket', 'basket_for_clothes', 'check', 'mouthwash', 'television', 'food_onion', 'doorjamb', 'controller', 'door', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control", "actions": ["[WALK] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'freezer', 'blow_dryer', 'bathroom', 'detergent', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'toaster', 'faucet', 'stereo', 'couch', 'pillow', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'bed', 'photoframe', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Washing Hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[TOUCH] <soap>', '[PUTOBJBACK] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'band-aids', 'walllamp', 'drawing', 'bills', 'orchid', 'floor', 'ceiling', 'character', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'plate', 'oil', 'mousepad', 'wall', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'doorjamb', 'door', 'napkin', 'table', 'teeth']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch TV in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'water_glass', 'pot', 'glue', 'walllamp', 'trashcan', 'knifeblock', 'ceiling', 'clothes_jacket', 'stovefan', 'toaster', 'mat', 'faucet', 'microphone', 'wall', 'cup', 'bed', 'food_food', 'check', 'water', 'food_donut', 'curtain', 'kitchen_counter', 'chessboard', 'cupboard', 'laundry_detergent', 'sink', 'spoon', 'bathroom_cabinet', 'dirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'cat', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'table_cloth', 'mat', 'plate', 'stereo', 'ceilinglamp', 'tvstand', 'electric_shaver', 'microphone', 'wall', 'powersocket', 'light_bulb', 'keys', 'doorjamb', 'pencil', 'laundry_detergent', 'door', 'clothes_scarf', 'sink', 'spoon', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I open the room door and press the light button on", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'drawing', 'orchid', 'knifeblock', 'bills', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'shower', 'stereo', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'television', 'food_rice', 'tray', 'curtain', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'home_office', 'table', 'light', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the sofa. Sit down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'jelly', 'chair', 'freezer', 'bathroom', 'food_carrot', 'knifeblock', 'bathroom_counter', 'keyboard', 'alcohol', 'clothes_jacket', 'stovefan', 'mat', 'toaster', 'stereo', 'picture', 'ceilinglamp', 'dining_room', 'bowl', 'wall', 'oven', 'powersocket', 'cup', 'photoframe', 'check', 'video_game_controller', 'razor', 'food_food', 'curtain', 'shoes']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "turn radio on, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'cat', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'character', 'desk', 'mat', 'deck_of_cards', 'iron', 'stereo', 'balanceball', 'shower', 'faucet', 'mousepad', 'nightstand', 'oven', 'television', 'tray', 'curtain', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'clothes_dress', 'sink', 'vase']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the hallway.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'freezer', 'floor', 'phone', 'character', 'thread', 'ceilinglamp', 'stereo', 'shower', 'window', 'mousepad', 'pillow', 'broom', 'wall', 'cup', 'photoframe', 'clothes_gloves', 'food_food', 'coffe_maker', 'razor', 'lighter', 'doorjamb', 'shoes', 'hairbrush', 'napkin', 'bathroom_cabinet', 'table', 'laser_pointer', 'clothes_shirt']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the radio; switch on the radio and listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'microwave', 'chair', 'band-aids', 'woman', 'needle', 'longboard', 'man', 'keyboard', 'envelope', 'toaster', 'mat', 'ceilinglamp', 'shower', 'faucet', 'stereo', 'nightstand', 'box', 'light_bulb', 'coffe_maker', 'curtain', 'towel_rack', 'shoes', 'dresser', 'hairbrush', 'door', 'vase', 'instrument_violin', 'creditcard']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'dog', 'walllamp', 'bathroom', 'drawing', 'knifeblock', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'computer', 'bookmark', 'ceilinglamp', 'tvstand', 'shower', 'window', 'foundation', 'couch', 'pillow', 'wall', 'razor', 'toothbrush_holder', 'tray', 'curtain', 'mouse', 'clothes_dress', 'sink', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "allow sofa to support weight in order to decrease gravity induced entropy", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'blender', 'crayon', 'walllamp', 'cat', 'drawing', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'bookmark', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'oven', 'cup', 'bed', 'clothes_gloves', 'newspaper', 'coffe_maker', 'curtain', 'hairbrush', 'food_steak', 'bag', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, find sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['jelly', 'trashcan', 'cpuscreen', 'clothes_underwear', 'ceiling', 'folder', 'desk', 'computer', 'stovefan', 'shower', 'mousepad', 'candle', 'clothes_hat', 'dining_room', 'newspaper', 'photoframe', 'food_food', 'video_game_controller', 'oven_mitts', 'curtain', 'pencil', 'kitchen_counter', 'dishrack', 'laundry_detergent', 'door', 'light', 'dishwasher', 'soap', 'tooth_paste', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.", "actions": ["[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'mop', 'toilet', 'freezer', 'walllamp', 'bathroom', 'cat', 'food_butter', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'iron', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'toy', 'tape', 'photoframe', 'television', 'food_rice', 'doorjamb', 'cards', 'door']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'ground_coffee', 'detergent', 'trashcan', 'orchid', 'drawing', 'bedroom', 'tea_bag', 'bookshelf', 'envelope', 'alcohol', 'computer', 'ceilinglamp', 'mat', 'shower', 'deck_of_cards', 'food_noodles', 'microphone', 'bowl', 'wall', 'bench', 'powersocket', 'photoframe', 'television', 'dresser', 'napkin', 'table', 'clothes_shirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'chair', 'nail_polish', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'ceilinglamp', 'balanceball', 'shower', 'mousepad', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'toothbrush_holder', 'doorjamb', 'mouse', 'light', 'vase']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['food_salt', 'beer', 'microwave', 'chair', 'drawing', 'hanger', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'bookmark', 'laptop', 'ceilinglamp', 'faucet', 'shower', 'mousepad', 'couch', 'food_peanut_butter', 'photoframe', 'newspaper', 'razor', 'tray', 'kitchen_counter', 'vacuum_cleaner', 'clothes_dress', 'door', 'bathroom_cabinet', 'tooth_paste']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to my bed room and sit on the chair and in front of me is my desktop computer and start writing an email.", "actions": ["[WALK] <bathroom>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'filing_cabinet', 'nail_polish', 'toothbrush', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'food_food', 'toothbrush_holder', 'tray', 'dresser', 'cupboard', 'controller', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'light', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I go into the bathroom, go to the sink, get toothbrush and toothpaste. Put toothpaste on the toothbrush, brush teeth, turn on faucet, rinse brush, and return toothbrush.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'tablelamp', 'toilet', 'walllamp', 'trashcan', 'drawing', 'orchid', 'bills', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'plate', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'photoframe', 'television', 'curtain', 'doorjamb', 'console', 'door', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to living room, sit on couch, get remote, switch on tv, watch tv", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'oven', 'cutting_board', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'dresser', 'food_steak', 'sink', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I sit at my computer desk and open the browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'bathroom', 'trashcan', 'dough', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'food_cake', 'computer', 'toaster', 'ceilinglamp', 'mat', 'deck_of_cards', 'mousepad', 'clothes_socks', 'pillow', 'wall', 'powersocket', 'bench', 'oven', 'food_food', 'doorjamb', 'mouse', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'food_salt', 'walllamp', 'detergent', 'trashcan', 'knifeblock', 'drawing', 'floor', 'stamp', 'table_cloth', 'plate', 'faucet', 'deck_of_cards', 'clothes_pants', 'wall', 'bench', 'powersocket', 'tape', 'photoframe', 'food_food', 'tray', 'pencil', 'kitchen_counter', 'cupboard', 'maindoor', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'coffee']\nTask:Bring dirty plate to sink\nresponse:\n{\n    \"thoughts\": Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[PUTBACK] <plate> <sink>']\n}\n", "input": "\nScene_Objects:['sheets', 'freezer', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'shower', 'stereo', 'mousepad', 'wall', 'powersocket', 'food_food', 'food_donut', 'tray', 'after_shave', 'doorjamb', 'cards', 'mouse', 'shoes', 'dresser', 'food_steak', 'door', 'bathroom_cabinet', 'light', 'clothes_skirt']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Go kitchen and make something for eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'freezer', 'cat', 'orchid', 'drawing', 'hanger', 'ceiling', 'dvd_player', 'desk', 'mail', 'piano_bench', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'food_orange', 'pillow', 'bench', 'bed', 'check', 'doorjamb', 'cd', 'controller', 'laundry_detergent', 'sink', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I go to the living room then go to the cat and touch it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'walllamp', 'bathroom', 'drawing', 'orchid', 'hanger', 'floor', 'phone', 'character', 'stamp', 'tea_bag', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'iron', 'brush', 'window', 'couch', 'food_orange', 'pillow', 'wall', 'bench', 'tray', 'curtain', 'after_shave', 'sink', 'bathroom_cabinet', 'table', 'home_office', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Sitting down and relaxing on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'drawing', 'hanger', 'floor', 'ceiling', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'foundation', 'oil', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'novel', 'television', 'curtain', 'after_shave', 'door', 'napkin', 'home_office', 'table']\nTask:Turn off TV\nresponse:\n{\n    \"thoughts\": Walk over to tv and press power button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'blender', 'food_salt', 'jelly', 'freezer', 'orchid', 'drawing', 'clothes_underwear', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'bookshelf', 'toaster', 'tvstand', 'shower', 'mousepad', 'clothes_pants', 'wall', 'powersocket', 'bed', 'food_food', 'tray', 'doorjamb', 'cards', 'controller', 'door', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I enter the bedroom and find the light. Then I turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'tablelamp', 'microwave', 'detergent', 'clothes_underwear', 'ceiling', 'character', 'desk', 'mail', 'table_cloth', 'bookmark', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'mousepad', 'candle', 'food_jam', 'food_orange', 'nightstand', 'powersocket', 'newspaper', 'coin', 'food_onion', 'pencil', 'dresser', 'home_office', 'dishwasher', 'tooth_paste', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'freezer', 'bathroom', 'detergent', 'food_butter', 'drawing', 'hanger', 'phone', 'bedroom', 'desk', 'toaster', 'brush', 'stereo', 'faucet', 'window', 'couch', 'pillow', 'dining_room', 'nightstand', 'powersocket', 'cup', 'coffe_maker', 'mouthwash', 'television', 'curtain', 'doorjamb', 'mouse', 'napkin', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'cat', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'ceilinglamp', 'brush', 'faucet', 'stereo', 'tvstand', 'towel', 'pillow', 'wall', 'pasta', 'powersocket', 'coffe_maker', 'oven_mitts', 'tray', 'curtain', 'dresser', 'cupboard', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'bathroom', 'food_carrot', 'knifeblock', 'hanger', 'bookshelf', 'computer', 'button', 'stovefan', 'couch', 'candle', 'dining_room', 'wall', 'bench', 'basket_for_clothes', 'newspaper', 'light_bulb', 'check', 'coffe_maker', 'television', 'after_shave', 'curtain', 'comb', 'pajamas', 'hair', 'door', 'sink', 'dishwasher', 'colander']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find the cat. Walk over to the cat. Touch the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'chair', 'freezer', 'ground_coffee', 'floor', 'ceiling', 'mail', 'food_bacon', 'toaster', 'deck_of_cards', 'faucet', 'window', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toy', 'light_bulb', 'coffe_maker', 'coffee_filter', 'water', 'curtain', 'after_shave', 'pencil', 'kitchen_counter', 'pajamas', 'sink', 'light']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[FIND] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'dog', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'envelope', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'pasta', 'nightstand', 'bench', 'photoframe', 'tray', 'mouse', 'door', 'table', 'soap']\nTask:Work\nresponse:\n", "output": {"thoughts": "go to home office. turn the computer on. begin working", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'door', 'home_office', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the remote. Pick up the remote. Find the sofa. Sit on the sofa. Look at the television. Aim the remote at the television. Press the power button. Flip the channels.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[PUSH] <remote_control>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'blender', 'crayon', 'walllamp', 'cat', 'drawing', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'bookmark', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'oven', 'cup', 'bed', 'clothes_gloves', 'newspaper', 'coffe_maker', 'curtain', 'hairbrush', 'food_steak', 'bag', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, find sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'chair', 'band-aids', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'laptop', 'toaster', 'foundation', 'pillow', 'wall', 'powersocket', 'bench', 'novel', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'napkin', 'light', 'soap', 'instrument_violin']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Grasp light switch, flip light switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[TURNTO] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['walllamp', 'dough', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'window', 'mousepad', 'food_noodles', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'pencil', 'hair', 'door', 'table', 'instrument_violin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the living room and sit on the couch. Turn on television using remote control. Use remote control buttons to navigate to desired channel. Watch television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'crayon', 'chair', 'cat', 'dough', 'milk', 'hanger', 'floor', 'keyboard', 'mail', 'table_cloth', 'stovefan', 'plate', 'faucet', 'tvstand', 'window', 'couch', 'form', 'dining_room', 'cutting_board', 'photoframe', 'check', 'wine', 'dresser', 'hairbrush', 'pajamas', 'conditioner', 'home_office', 'bathroom_cabinet', 'table']\nTask:Give milk to cat\nresponse:\n{\n    \"thoughts\": Grab milk and pour it into the plate so that the cat can drink,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[POINTAT] <cat>', '[PULL] <cat>', '[FIND] <milk>', '[GRAB] <milk>', '[FIND] <plate>', '[POUR] <milk> <plate>', '[TOUCH] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['blender', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'basket_for_clothes', 'cup', 'television', 'food_rice', 'curtain', 'doorjamb', 'diary', 'shoes', 'door', 'spoon', 'bathroom_cabinet', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room and sit on the sofa to watch tv", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'sponge', 'scissors', 'food_salt', 'trashcan', 'drawing', 'needle', 'hanger', 'longboard', 'floor', 'character', 'stamp', 'keyboard', 'stovefan', 'ceilinglamp', 'deck_of_cards', 'window', 'couch', 'dining_room', 'wall', 'photoframe', 'television', 'curtain', 'rag', 'pencil', 'shoes', 'clothes_dress', 'conditioner', 'sink', 'table']\nTask:Text friends while sitting on couch\nresponse:\n{\n    \"thoughts\": Text while sitting,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[TOUCH] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'toaster', 'deck_of_cards', 'window', 'tvstand', 'mousepad', 'pillow', 'wall', 'pasta', 'light_bulb', 'check', 'mouthwash', 'wooden_spoon', 'mouse', 'door', 'table', 'coffee']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Turn on computer, pull up browser, browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'ceilinglamp', 'tvstand', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'clothes_gloves', 'novel', 'tray', 'doorjamb', 'kitchen_counter', 'door', 'table', 'light', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'tablelamp', 'microwave', 'drawing', 'knife', 'knifeblock', 'hanger', 'bathroom_counter', 'phone', 'stovefan', 'laptop', 'faucet', 'foundation', 'couch', 'food_orange', 'pillow', 'bowl', 'bench', 'oven', 'basket_for_clothes', 'cup', 'novel', 'razor', 'centerpiece', 'doorjamb', 'shaving_cream', 'mouse', 'door', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the living room, walk to the lamp, switch on the lamp.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'alcohol', 'toaster', 'tvstand', 'shower', 'closetdrawer', 'nightstand', 'wall_clock', 'bench', 'powersocket', 'food_food', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'sink', 'home_office', 'table', 'light', 'coffee']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the hallway.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dining_room>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sheets', 'sponge', 'chair', 'freezer', 'trashcan', 'ceiling', 'stamp', 'desk', 'keyboard', 'table_cloth', 'stovefan', 'ceilinglamp', 'shower', 'iron', 'electric_shaver', 'mousepad', 'clothes_socks', 'powersocket', 'oven', 'tray', 'kitchen_counter', 'dishrack', 'hair', 'cupboard', 'door', 'hairdryer', 'bathroom_cabinet', 'instrument_guitar', 'coffee']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room ,find radio ,switch on radio ,listen to music 4", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'microwave', 'freezer', 'ground_coffee', 'dough', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'bedroom', 'tea_bag', 'toaster', 'faucet', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toy', 'coffe_maker', 'television', 'coffee_filter', 'water', 'mouse', 'kitchen_counter', 'sink', 'table']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['glue', 'food_bread', 'mop', 'freezer', 'trashcan', 'knifeblock', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'stovefan', 'mat', 'faucet', 'pillow', 'wall', 'powersocket', 'bed', 'food_food', 'check', 'mouthwash', 'after_shave', 'doorjamb', 'chessboard', 'dresser', 'cupboard', 'hair', 'door', 'instrument_guitar', 'light', 'creditcard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into living room. Find wall. Find light button. Switch on light button.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I need to write an e-mail to my boss about the meeting at work tomorrow.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[TYPE] <keyboard>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'knifeblock', 'floor', 'ceiling', 'purse', 'computer', 'table_cloth', 'stovefan', 'mat', 'faucet', 'window', 'bowl', 'wall', 'powersocket', 'cutting_board', 'television', 'razor', 'tray', 'after_shave', 'doorjamb', 'shoes', 'comb', 'dresser', 'door', 'home_office', 'light', 'dishwasher', 'soap', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk in to the living room. I find the light switch. I turn on the lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'measuring_cup', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'keyboard', 'mail', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'window', 'couch', 'candle', 'pillow', 'bowl', 'wall', 'powersocket', 'television', 'curtain', 'after_shave', 'doorjamb', 'chessboard', 'door', 'food_steak', 'bathroom_cabinet', 'table', 'board_game']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'powersocket', 'clothes_gloves', 'television', 'razor', 'curtain', 'doorjamb', 'console', 'mouse', 'door', 'bag', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk in living room. I find TV remote. I grab TV remote. I sit on sofa. I switch on TV. I switch channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'tea_bag', 'computer', 'food_bacon', 'tvstand', 'iron', 'window', 'electric_shaver', 'mousepad', 'dining_room', 'wall', 'nightstand', 'wall_clock', 'photoframe', 'towel_rack', 'mouse', 'table']\nTask:Check email\nresponse:\n{\n    \"thoughts\": Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[WALK] <chair>', '[SIT] <chair>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'cpuscreen', 'hanger', 'floor', 'keyboard', 'purse', 'bookshelf', 'food_bacon', 'button', 'toaster', 'laptop', 'stereo', 'couch', 'wall', 'bench', 'basket_for_clothes', 'toothbrush_holder', 'curtain', 'rag', 'diary', 'cards', 'cupboard', 'controller', 'laundry_detergent', 'soap', 'tooth_paste', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'tvstand', 'shower', 'closetdrawer', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'powersocket', 'cup', 'bed', 'coffe_maker', 'food_food', 'doorjamb', 'towel_rack', 'pencil', 'headset', 'chessboard', 'dresser', 'hairbrush', 'door', 'clothes_scarf', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go into my room, find the light,  then turn on the lights.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'tablelamp', 'microwave', 'detergent', 'clothes_underwear', 'ceiling', 'character', 'desk', 'mail', 'table_cloth', 'bookmark', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'mousepad', 'candle', 'food_jam', 'food_orange', 'nightstand', 'powersocket', 'newspaper', 'coin', 'food_onion', 'pencil', 'dresser', 'home_office', 'dishwasher', 'tooth_paste', 'coffee']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "The task is browse and surf on the internet regarding a particular topic", "actions": ["[WALK] <bedroom>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[WATCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_salt', 'toilet', 'detergent', 'cpuscreen', 'food_egg', 'bathroom_counter', 'floor', 'folder', 'desk', 'envelope', 'clothes_jacket', 'mat', 'plate', 'stereo', 'deck_of_cards', 'shower', 'couch', 'television', 'oven_mitts', 'curtain', 'shaving_cream', 'chessboard', 'cupboard', 'conditioner', 'light', 'table', 'soap', 'clothes_shirt']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Get a broom or duster with long handle. Dust blades of ceiling fan.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <duster>', '[FIND] <duster>', '[GRAB] <duster>', '[PUTOBJBACK] <duster>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'stamp', 'keyboard', 'computer', 'bookmark', 'ceilinglamp', 'mat', 'mousepad', 'clothes_pants', 'wall', 'cutting_board', 'food_cheese', 'keys', 'food_food', 'video_game_controller', 'toothbrush_holder', 'mouse', 'pencil', 'door', 'sink', 'table', 'light', 'dirt']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go to home office, turn on the computer, open the internet and browse around the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'bathroom', 'knifeblock', 'bathroom_counter', 'shampoo', 'phone', 'mail', 'clothes_jacket', 'stovefan', 'ceilinglamp', 'window', 'clothes_socks', 'couch', 'clothes_pants', 'wall', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'television', 'mouse', 'shoes', 'chessboard', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'dishwasher', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Open drawer. Pull out pants and put them on. Take a shirt out and put it on. Put on socks and shoes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTON] <clothes_socks>', '[CLOSE] <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'toilet', 'food_butter', 'orchid', 'bathroom_counter', 'bedroom', 'tea_bag', 'bookshelf', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'window', 'food_noodles', 'couch', 'clothes_hat', 'pillow', 'bench', 'powersocket', 'basket_for_clothes', 'television', 'oven_mitts', 'kitchen_counter', 'shoes', 'pajamas', 'door', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_ice_cream', 'beer', 'microwave', 'freezer', 'walllamp', 'floor', 'ceiling', 'phone', 'dvd_player', 'stovefan', 'ceilinglamp', 'toaster', 'mousepad', 'wall', 'wall_clock', 'cup', 'bed', 'food_food', 'coin', 'food_rice', 'water', 'pencil', 'shoes', 'controller', 'napkin', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sponge', 'measuring_cup', 'walllamp', 'nail_polish', 'drawing', 'orchid', 'floor', 'ceiling', 'dvd_player', 'keyboard', 'computer', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'window', 'couch', 'pillow', 'broom', 'wall', 'powersocket', 'oven', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'mouse', 'bag', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I sit on the sofa. I put my legs on the sofa and recline.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'blow_dryer', 'walllamp', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'faucet', 'picture', 'iron', 'mousepad', 'bowl', 'wall', 'powersocket', 'bench', 'photoframe', 'toilet_paper', 'coffe_maker', 'toothbrush_holder', 'tray', 'doorjamb', 'dishrack', 'sink', 'bathroom_cabinet', 'home_office', 'light']\nTask:Get toilet paper\nresponse:\n{\n    \"thoughts\": There are extra toilet papers in the bottom cabinet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <toilet_paper>', '[GRAB] <toilet_paper>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'trashcan', 'dough', 'food_carrot', 'knife', 'drawing', 'clothes_underwear', 'hanger', 'dvd_player', 'character', 'bookmark', 'toaster', 'mat', 'faucet', 'oven', 'tape', 'food_food', 'television', 'coffee_filter', 'ice', 'wooden_spoon', 'tray', 'lighter', 'chessboard', 'spoon', 'home_office', 'table', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I hear the phone ringing in the living room. I walk to the living room and go to the phone. I pick it up and answer the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>", "[SWITCHON] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'bathroom', 'trashcan', 'food_butter', 'drawing', 'floor', 'ceiling', 'phone', 'mat', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'food_food', 'check', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'door', 'sink', 'spoon', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['freezer', 'bathroom', 'detergent', 'cat', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'mousepad', 'form', 'wall', 'powersocket', 'photoframe', 'food_food', 'check', 'coffee_filter', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'controller', 'door', 'light', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'walllamp', 'nail_polish', 'dough', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'cards', 'dishrack', 'door', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'crayon', 'walllamp', 'drawing', 'knifeblock', 'orchid', 'floor', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'window', 'couch', 'food_orange', 'pillow', 'dining_room', 'wall', 'powersocket', 'bench', 'cutting_board', 'curtain', 'cards', 'dresser', 'bag', 'table', 'light', 'colander', 'clothes_shirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room, turn to the sofa. Sit on the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'console', 'dresser', 'cupboard', 'home_office', 'table', 'clothes_skirt', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'walllamp', 'food_butter', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'mousepad', 'wall', 'powersocket', 'photoframe', 'newspaper', 'ice', 'curtain', 'mouse', 'headset', 'hairbrush', 'door', 'bathroom_cabinet', 'table', 'clothes_shirt']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "walk into office, sit in chair, turn on computer, write email", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'detergent', 'knifeblock', 'phone', 'bookshelf', 'computer', 'mop_bucket', 'brush', 'clothes_socks', 'mousepad', 'dining_room', 'nightstand', 'basket_for_clothes', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'bathtub', 'sink', 'home_office', 'napkin', 'dishwasher', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'mop', 'microwave', 'ground_coffee', 'walllamp', 'cat', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'toaster', 'brush', 'tvstand', 'shower', 'faucet', 'food_jam', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'mouthwash', 'coffee_filter', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'dishwasher', 'scrabble']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Make coffee in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_pizza', 'walllamp', 'orchid', 'needle', 'cpuscreen', 'bathroom_counter', 'shampoo', 'desk', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'shower', 'faucet', 'couch', 'dining_room', 'cup', 'bed', 'tray', 'curtain', 'wooden_spoon', 'pencil', 'door', 'bathroom_cabinet', 'light', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office. I find the telephone. I pick up the telephone handset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'filing_cabinet', 'dog', 'bathroom', 'knifeblock', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'character', 'desk', 'bedroom', 'clothes_jacket', 'mat', 'tvstand', 'pillow', 'towel', 'wall', 'cup', 'food_donut', 'curtain', 'after_shave', 'dishrack', 'cupboard', 'controller', 'door', 'dishwasher']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'tvstand', 'faucet', 'plate', 'mousepad', 'wall', 'powersocket', 'bench', 'oven', 'cutting_board', 'tape', 'cup', 'doorjamb', 'sink', 'light', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk over and pick up the phone,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'freezer', 'walllamp', 'bathroom', 'nail_polish', 'detergent', 'trashcan', 'drawing', 'floor', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'table_cloth', 'stereo', 'mousepad', 'broom', 'wall', 'nightstand', 'powersocket', 'oven', 'television', 'rag', 'mouse', 'headset', 'door', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to my bedroom.  I locate my stereo and turn it on.  I put on my headphones and sit to listen.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[WALK] <stereo>", "[SWITCHON] <stereo>", "[FIND] <headset>", "[GRAB] <headset>", "[PUTON] <headset>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'bookshelf', 'keyboard', 'brush', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'stovefan', 'shoe-shine_kit', 'pillow', 'wall', 'powersocket', 'photoframe', 'novel', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'bathroom_cabinet', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit on the couch. Turn on TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'band-aids', 'detergent', 'toothbrush', 'drawing', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'purse', 'stovefan', 'mat', 'tvstand', 'faucet', 'brush', 'window', 'oil', 'couch', 'bowl', 'wall', 'powersocket', 'coffee_pot', 'pencil', 'food_steak', 'sink', 'bag', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I walk in to the bathroom. I turn on the faucet. I find the toothbrush and toothpaste. I grab the toothbrush and toothpaste. I rinse off the toothbrush. I add toothpaste to the toothbrush. I put back the toothpaste. I wash my teeth with the toothbrush. I rinse off the toothbrush. I put the toothbrush back. I rinse my mouth out. I turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[RINSE] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>", "[PUTOBJBACK] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>", "[RINSE] <toothbrush>", "[PUTOBJBACK] <toothbrush>", "[RINSE] <teeth>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'detergent', 'toothbrush', 'drawing', 'cpuscreen', 'food_egg', 'phone', 'character', 'computer', 'piano_bench', 'shoe-shine_kit', 'faucet', 'food_noodles', 'mousepad', 'clothes_hat', 'towel', 'broom', 'cutting_board', 'shelf', 'keys', 'novel', 'food_food', 'check', 'wooden_spoon', 'towel_rack', 'sink', 'table', 'light', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'tablelamp', 'beer', 'bathroom', 'food_butter', 'food_carrot', 'milk', 'bathroom_counter', 'bedroom', 'thread', 'table_cloth', 'bookmark', 'ceilinglamp', 'shower', 'mousepad', 'nightstand', 'powersocket', 'box', 'photoframe', 'razor', 'wine', 'dresser', 'clothes_scarf', 'sink', 'home_office', 'light', 'colander', 'laser_pointer', 'dirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'freezer', 'trashcan', 'drawing', 'bathroom_counter', 'ceiling', 'phone', 'thread', 'computer', 'stovefan', 'faucet', 'shower', 'food_jam', 'dining_room', 'wall', 'powersocket', 'tape', 'coffe_maker', 'food_food', 'mirror', 'doorjamb', 'mouse', 'paper_towel', 'dresser', 'cupboard', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'home_office', 'dirt']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": Spray Windex or glass and mirror cleaner onto mirror. Use paper towel to wipe off cleaner until mirror is dry and has no streaks.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[WIPE] <mirror>', '[FIND] <trashcan>', '[PUTBACK] <paper_towel> <trashcan>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'measuring_cup', 'microwave', 'detergent', 'trashcan', 'orchid', 'bills', 'knife', 'bathroom_counter', 'floor', 'character', 'keyboard', 'mat', 'mousepad', 'nightstand', 'pasta', 'tape', 'novel', 'food_food', 'check', 'television', 'curtain', 'coffee_pot', 'laundry_detergent', 'clothes_scarf', 'bathroom_cabinet', 'colander', 'instrument_violin', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to bedroom and grab dirty pants and take to the washing machine in entrance hall. Put clothes in washing machine and then put in laundry detergent.", "actions": ["[WALK] <bedroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'woman', 'toothbrush', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'stamp', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'child', 'box', 'photoframe', 'television', 'oven_mitts', 'doorjamb', 'cd', 'dresser', 'table']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Sitting and watching  a movie with family,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TURNTO] <child>', '[LOOKAT] <child>', '[GREET] <child>', '[FIND] <woman>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>', '[WALK] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'blender', 'walllamp', 'cat', 'drawing', 'orchid', 'needle', 'bathroom_counter', 'floor', 'bedroom', 'desk', 'ceilinglamp', 'laptop', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'bench', 'oven', 'newspaper', 'check', 'tray', 'curtain', 'doorjamb', 'door', 'bathroom_cabinet', 'table', 'tooth_paste', 'coffee']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I'm tired and want to relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'folder', 'bedroom', 'food_cake', 'computer', 'shower', 'iron', 'window', 'pillow', 'wall', 'nightstand', 'powersocket', 'tape', 'bed', 'television', 'wine', 'doorjamb', 'towel_rack', 'dresser', 'home_office', 'light', 'bathroom_cabinet', 'dishwasher']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the phone. I pick up the phone. I hold it to my ear.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'food_salt', 'microwave', 'blow_dryer', 'detergent', 'knife', 'orchid', 'bathroom_counter', 'shampoo', 'keyboard', 'bookshelf', 'mail', 'stovefan', 'ceilinglamp', 'toaster', 'mat', 'faucet', 'food_peanut_butter', 'dining_room', 'nightstand', 'oven', 'cup', 'light_bulb', 'food_food', 'television', 'food_onion', 'pencil', 'napkin', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'food_butter', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'mop_bucket', 'ceilinglamp', 'tvstand', 'iron', 'window', 'mousepad', 'microphone', 'towel', 'dining_room', 'bowl', 'wall', 'powersocket', 'bench', 'cup', 'novel', 'television', 'coin', 'rag', 'doorjamb', 'headset', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk in to the living room. I find the light switch. I turn on the lights.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'measuring_cup', 'microwave', 'walllamp', 'trashcan', 'floor', 'bedroom', 'computer', 'toaster', 'stovefan', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'candle', 'pillow', 'wall', 'nightstand', 'bench', 'oven', 'cup', 'newspaper', 'coffe_maker', 'tray', 'kitchen_counter', 'cupboard', 'home_office', 'dishwasher', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the sink. Take dishes out of sink. Open dishwasher and place dishes in dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_bread', 'chair', 'freezer', 'walllamp', 'detergent', 'trashcan', 'drawing', 'desk', 'keyboard', 'bookshelf', 'bookmark', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'pillow', 'bowl', 'wall', 'box', 'coffe_maker', 'razor', 'television', 'curtain', 'doorjamb', 'towel_rack', 'door', 'sink', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Pour water into a glass and drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['freezer', 'ground_coffee', 'bathroom', 'drawing', 'floor', 'ceiling', 'bookshelf', 'mat', 'ceilinglamp', 'stereo', 'mousepad', 'couch', 'clothes_pants', 'dining_room', 'wall', 'powersocket', 'box', 'bed', 'check', 'television', 'wooden_spoon', 'after_shave', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'food_steak', 'napkin', 'light', 'board_game']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on the light", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'envelope', 'computer', 'button', 'food_bacon', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'shoe-shine_kit', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'photoframe', 'television', 'doorjamb', 'dresser', 'dirt']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Pickup remote control for the living room TV and push the change channel button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'cpuscreen', 'clothes_underwear', 'hanger', 'shampoo', 'bookshelf', 'alcohol', 'table_cloth', 'faucet', 'shower', 'clothes_socks', 'clothes_pants', 'wall', 'wall_clock', 'oven', 'nightstand', 'basket_for_clothes', 'newspaper', 'clothes_gloves', 'tray', 'doorjamb', 'lighter', 'cards', 'clothes_dress', 'laundry_detergent', 'napkin', 'clothes_skirt', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom, take the dirt clothes and put into the washing machine, then put the soap, set the cleaning program and switch on the machine", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_skirt>", "[GRAB] <clothes_skirt>", "[PUTBACK] <clothes_skirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[CLOSE] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'cat', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'mail', 'alcohol', 'mat', 'faucet', 'clothes_pants', 'pillow', 'wall', 'wall_clock', 'oven', 'pasta', 'bench', 'coffee_filter', 'ice', 'paper_towel', 'conditioner', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk to the bathroom and find the sink. I grab the soap and rub some on my hands. I turn the faucet on and put my hands in the water. I rinse my hands and then turn off the faucet. Then I dry my hands on a paper towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <hands_both>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'faucet', 'picture', 'window', 'mousepad', 'oil', 'clothes_hat', 'wall', 'keys', 'food_food', 'coin', 'food_onion', 'mouse', 'cupboard', 'laundry_detergent', 'door', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I walk into the home office and walk to the desk. I pull out the chair and sit down. I turn on the computer. After the computer is turned on I click on my email program.  When the program is loaded I write an email.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'measuring_cup', 'dog', 'freezer', 'walllamp', 'orchid', 'bathroom_counter', 'keyboard', 'bookshelf', 'closetdrawer', 'pillow', 'dining_room', 'powersocket', 'oven', 'newspaper', 'clothes_gloves', 'bed', 'check', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'cd', 'cupboard', 'door', 'napkin', 'home_office', 'bathroom_cabinet', 'dishwasher', 'scrabble']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Read a book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'floor', 'character', 'desk', 'bedroom', 'keyboard', 'thread', 'toaster', 'plate', 'couch', 'wall', 'light_bulb', 'novel', 'coffe_maker', 'food_rice', 'tray', 'food_onion', 'after_shave', 'curtain', 'kettle', 'kitchen_counter', 'shoes', 'cupboard', 'conditioner', 'door', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'chair', 'bathroom', 'food_carrot', 'woman', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'man', 'keyboard', 'mat', 'faucet', 'window', 'foundation', 'nightstand', 'powersocket', 'cup', 'coffe_maker', 'food_food', 'television', 'tray', 'towel_rack', 'cd', 'dresser', 'sink', 'clothes_scarf', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Stand next to the door. When the door opens from the outside, greet the guests.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <woman>', '[FIND] <woman>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>', '[FIND] <man>', '[TURNTO] <man>', '[LOOKAT] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'measuring_cup', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'keyboard', 'mail', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'window', 'couch', 'candle', 'pillow', 'bowl', 'wall', 'powersocket', 'television', 'curtain', 'after_shave', 'doorjamb', 'chessboard', 'door', 'food_steak', 'bathroom_cabinet', 'table', 'board_game']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tea', 'ground_coffee', 'walllamp', 'hanger', 'longboard', 'floor', 'ceiling', 'button', 'toaster', 'faucet', 'mousepad', 'candle', 'wall', 'nightstand', 'powersocket', 'keys', 'novel', 'coffe_maker', 'clothes_gloves', 'coffee_filter', 'water', 'ice', 'tray', 'food_onion', 'cupboard', 'door', 'light', 'dishwasher', 'soap']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'walllamp', 'bathroom', 'toothbrush', 'drawing', 'orchid', 'floor', 'desk', 'computer', 'ceilinglamp', 'mat', 'deck_of_cards', 'stereo', 'window', 'couch', 'food_jam', 'pillow', 'wall', 'light_bulb', 'food_donut', 'tray', 'curtain', 'shoes', 'dishrack', 'sink', 'spoon', 'table', 'dishwasher', 'coffee']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'tea_bag', 'bookshelf', 'food_bacon', 'toaster', 'plate', 'tvstand', 'dish_soap', 'mousepad', 'couch', 'wall', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'food_food', 'toothbrush_holder', 'ice', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'clothes_dress', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'oil', 'mousepad', 'food_peanut_butter', 'wall', 'oven', 'powersocket', 'food_food', 'water', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I need to go to Kitchen.  Get a cup.  Fill cup.  Drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'microwave', 'chair', 'wallshelf', 'hanger', 'floor', 'phone', 'food_bacon', 'piano_bench', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'powersocket', 'television', 'coin', 'food_donut', 'ice', 'rag', 'pencil', 'shoes', 'dresser', 'clothes_dress', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Hang up jacket\nresponse:\n{\n    \"thoughts\": I put clothes on hangers and hang them in the closet,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[FIND] <hanger>', '[OPEN] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['glue', 'microwave', 'walllamp', 'ground_coffee', 'trashcan', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'iron', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'door', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the sofa and watch TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'scissors', 'walllamp', 'detergent', 'drawing', 'clothes_underwear', 'floor', 'phone', 'character', 'button', 'toaster', 'plate', 'tvstand', 'dish_soap', 'oil', 'candle', 'wall', 'oven', 'tape', 'box', 'coffe_maker', 'food_food', 'check', 'tray', 'centerpiece', 'controller', 'light', 'dishwasher', 'soap']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": walk into kitchen, find dishwasher, put plates in dishwasher, put soap in dishwasher, turn on dishwasher,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'ceiling', 'bookshelf', 'envelope', 'toaster', 'mat', 'food_orange', 'pillow', 'wall', 'wall_clock', 'powersocket', 'cutting_board', 'bed', 'oven_mitts', 'coin', 'doorjamb', 'lighter', 'pencil', 'kitchen_counter', 'dresser', 'pajamas', 'door', 'home_office', 'light', 'colander']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk towards to the bedroom, and I open the door of the bedroom, and I switch on the lights inside the bedroom, the switch is beside the door.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'juice', 'blow_dryer', 'walllamp', 'detergent', 'floor', 'ceiling', 'bedroom', 'keyboard', 'purse', 'tvstand', 'deck_of_cards', 'clothes_pants', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'keys', 'tray', 'kettle', 'shoes', 'hair', 'sink', 'home_office', 'table', 'dishwasher', 'soap', 'board_game', 'dirt']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['sponge', 'dry_pasta', 'beer', 'microwave', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'drawing', 'knifeblock', 'floor', 'phone', 'bedroom', 'food_bacon', 'toaster', 'ceilinglamp', 'window', 'dish_soap', 'candle', 'pillow', 'bowl', 'wall', 'bench', 'newspaper', 'coffe_maker', 'kettle', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the dishwasher and put dish soap. I close the dishwasher and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'crayon', 'drawing', 'hanger', 'floor', 'character', 'bedroom', 'mat', 'window', 'towel', 'pillow', 'wall', 'nightstand', 'cutting_board', 'cup', 'bed', 'toilet_paper', 'food_donut', 'curtain', 'doorjamb', 'mouse', 'shoes', 'chessboard', 'controller', 'laundry_detergent', 'door', 'table', 'vase']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go t the bedroom and go to sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'food_salt', 'microwave', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'mat', 'dustpan', 'food_orange', 'wall', 'powersocket', 'oven', 'photoframe', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'pajamas', 'cupboard', 'clothes_dress', 'door', 'light', 'dishwasher', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the Living Room. Switch on 1 light. Walk to dining room. Switch on 1 light. Walk to Entrance hall. Switch on 1 light.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'toilet', 'chair', 'detergent', 'bathroom_counter', 'ceiling', 'character', 'desk', 'laptop', 'tvstand', 'faucet', 'electric_shaver', 'mousepad', 'couch', 'food_orange', 'bench', 'powersocket', 'cup', 'bed', 'novel', 'razor', 'oven_mitts', 'curtain', 'after_shave', 'cd', 'headset', 'dresser', 'pajamas', 'bathroom_cabinet', 'tooth_paste']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[WALK] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mail', 'stovefan', 'ceilinglamp', 'mat', 'mousepad', 'dustpan', 'wall', 'bed', 'coffe_maker', 'lighter', 'coffee_pot', 'mouse', 'hair', 'door', 'napkin', 'home_office', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go find a computer, turn it on, click google, and browse the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'walllamp', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'faucet', 'shower', 'towel', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'board_game', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, open the faucet rinse the hands, wash the hands with soap and rinse again, then wipe with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mop_bucket', 'ceilinglamp', 'mat', 'electric_shaver', 'mousepad', 'wall', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'mouse', 'pencil', 'controller', 'door', 'table', 'teeth']\nTask:Work\nresponse:\n", "output": {"thoughts": "walk to home office ,find computer ,switch on computer and work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office']\nTask:Check email\nresponse:\n{\n    \"thoughts\": Turn on the computer.  Wait for it to load.  Get online.  Go to the email service.  Open the email.  Read the email.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mail>', '[GRAB] <mail>', '[READ] <mail>']\n}\n", "input": "\nScene_Objects:['blender', 'food_ice_cream', 'walllamp', 'bathroom', 'knifeblock', 'cpuscreen', 'character', 'envelope', 'stovefan', 'stereo', 'shower', 'window', 'towel', 'dining_room', 'broom', 'nightstand', 'cup', 'basket_for_clothes', 'check', 'food_rice', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'chessboard', 'hairbrush', 'hair', 'dresser', 'light', 'shoe_rack']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <address_book>", "[FIND] <address_book>", "[GRAB] <address_book>", "[READ] <address_book>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'folder', 'character', 'desk', 'faucet', 'stereo', 'mousepad', 'food_orange', 'wall', 'oven', 'photoframe', 'oven_mitts', 'water', 'drinking_glass', 'wooden_spoon', 'towel_rack', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'spoon', 'board_game']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Go to kitchen, open cabinet, get glass, close cabinet, put glass under faucet, turn on water, turn off water when glass has 85% water, lift glass to mouth, open mouth and put on glass, tilt glass up and swallow water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[TURNTO] <kitchen_cabinet>', '[LOOKAT] <kitchen_cabinet>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[FIND] <sink>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[SWITCHOFF] <faucet>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'tea', 'walllamp', 'ground_coffee', 'detergent', 'trashcan', 'milk', 'drawing', 'knifeblock', 'orchid', 'floor', 'phone', 'bookshelf', 'stovefan', 'ceilinglamp', 'deck_of_cards', 'shower', 'window', 'couch', 'dustpan', 'pillow', 'dining_room', 'wall', 'nightstand', 'toy', 'photoframe', 'food_food', 'curtain', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I bend down. I sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'shampoo', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to my office. I switch the computer on and sit on the office chair. I look at the computer screen to see what needs to be done. I use my mouse to navigate to the relevant directory, then release it. I type code with my keyboard. When I'm done, I look at the computer again to evaluate what I've done. Then I switch off the computer using the mouse and put back the mouse.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[GRAB] <mouse>', '[SWITCHOFF] <computer>', '[PUTOBJBACK] <mouse>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'detergent', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'cutting_board', 'bed', 'television', 'curtain', 'doorjamb', 'hairbrush', 'controller', 'door', 'table', 'dishwasher', 'instrument_violin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to the living room. find the remote. sit on the couch. turn on the tv.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'mop', 'bathroom', 'knife', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'desk', 'food_cake', 'window', 'electric_shaver', 'microphone', 'food_jam', 'dining_room', 'wall', 'photoframe', 'food_food', 'television', 'curtain', 'mouse', 'kitchen_counter', 'cd', 'chessboard', 'dresser', 'hairbrush', 'clothes_scarf', 'bathroom_cabinet', 'dishwasher', 'instrument_violin']\nTask:Put on coat\nresponse:\n{\n    \"thoughts\": I remove my scarf from a coat peg in the entrance hall.  I wrap the scarf around my neck,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <hanger>', '[TURNTO] <hanger>', '[LOOKAT] <hanger>', '[FIND] <clothes_scarf>', '[GRAB] <clothes_scarf>', '[PUTON] <clothes_scarf>']\n}\n", "input": "\nScene_Objects:['pot', 'measuring_cup', 'microwave', 'chair', 'food_butter', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'desk', 'food_bacon', 'mat', 'tvstand', 'form', 'pillow', 'dining_room', 'wall', 'bench', 'cup', 'light_bulb', 'television', 'kettle', 'kitchen_counter', 'cd', 'cupboard', 'clothes_dress', 'door', 'bathroom_cabinet']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone. .I reach my arm out. I open my hand. I place my hand on the phone. I close my fingers around the phone. I pick up the phone.", "actions": ["[WALK] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'oven', 'photoframe', 'television', 'doorjamb', 'headset', 'dresser', 'bathroom_cabinet', 'soap']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I find the remote then I push the button to switch it to the tv channel i want to watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sheets', 'filing_cabinet', 'tablelamp', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'kitchen_cabinet', 'ceiling', 'bookshelf', 'piano_bench', 'plate', 'shower', 'window', 'couch', 'dining_room', 'pasta', 'oven', 'bench', 'television', 'water', 'drinking_glass', 'curtain', 'doorjamb', 'clothes_scarf', 'napkin', 'light', 'dishwasher', 'colander']\nTask:Drink\nresponse:\n", "output": {"thoughts": "You walk to the kitchen and open the kitchen cabinet and grab a drinking glass. You find the water and pour the water into the drinking glass. You drink the water from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'food_kiwi', 'wall', 'powersocket', 'clothes_gloves', 'food_food', 'coffe_maker', 'doorjamb', 'coffee_pot', 'headset', 'door', 'hairdryer', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk towards to the bedroom, and I open the door of the bedroom, and I switch on the lights inside the bedroom, the switch is beside the door.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'food_pizza', 'food_salt', 'microwave', 'floor', 'phone', 'ceiling', 'bedroom', 'computer', 'table_cloth', 'stovefan', 'shower', 'stereo', 'couch', 'clothes_pants', 'nightstand', 'oven', 'cup', 'photoframe', 'novel', 'television', 'food_donut', 'ice', 'doorjamb', 'mouse', 'light', 'colander', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room. Turn on radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'drawing', 'needle', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'dustpan', 'dining_room', 'wall', 'nightstand', 'powersocket', 'food_food', 'toilet_paper', 'doorjamb', 'coffee_pot', 'mouse', 'kettle', 'dresser', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk into living room, walk to light, turn on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'phone', 'piano_bench', 'toaster', 'mat', 'faucet', 'shower', 'window', 'dining_room', 'wall', 'bench', 'hands_both', 'bed', 'television', 'oven_mitts', 'water', 'curtain', 'paper_towel', 'hairdryer', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. I turn on the water. I put soap on my hands. I rinse my hands. I turn off the water. I dry my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <water>", "[TURNTO] <water>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[TURNTO] <water>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'tea', 'drawing', 'wallshelf', 'hanger', 'ceiling', 'tea_bag', 'bookshelf', 'stovefan', 'toaster', 'mat', 'faucet', 'floor_lamp', 'mousepad', 'nightstand', 'box', 'bed', 'photoframe', 'novel', 'food_food', 'coffee_filter', 'food_donut', 'wine', 'after_shave', 'curtain', 'doorjamb', 'clothes_dress', 'conditioner', 'sink']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Turn a light on and read a book in the bedroom.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_pizza', 'mop', 'cat', 'knife', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'dvd_player', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'faucet', 'stereo', 'window', 'couch', 'towel', 'wall', 'novel', 'tray', 'doorjamb', 'hairbrush', 'sink', 'bathroom_cabinet', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'chair', 'bathroom', 'food_butter', 'phone', 'ceiling', 'bedroom', 'tea_bag', 'laptop', 'ceilinglamp', 'shower', 'iron', 'clothes_socks', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'powersocket', 'standingmirror', 'photoframe', 'toothbrush_holder', 'food_donut', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'controller', 'door', 'table']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['tea', 'toilet', 'freezer', 'band-aids', 'knifeblock', 'food_egg', 'shampoo', 'phone', 'bookshelf', 'purse', 'thread', 'deck_of_cards', 'shower', 'picture', 'clothes_socks', 'oven', 'food_cheese', 'novel', 'food_food', 'check', 'coffe_maker', 'tray', 'after_shave', 'curtain', 'doorjamb', 'cupboard', 'controller', 'bag', 'table', 'scrabble']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Pour water into a glass and drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'microwave', 'toilet', 'cat', 'wallshelf', 'bathroom_counter', 'floor', 'dvd_player', 'bookshelf', 'stovefan', 'ceilinglamp', 'toaster', 'mat', 'shower', 'faucet', 'tvstand', 'couch', 'towel', 'wall', 'bench', 'cutting_board', 'bed', 'razor', 'television', 'curtain', 'kitchen_counter', 'dresser', 'home_office', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Pet Cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'blender', 'beer', 'freezer', 'band-aids', 'walllamp', 'orchid', 'food_egg', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'button', 'ceilinglamp', 'mousepad', 'food_peanut_butter', 'wall', 'powersocket', 'food_food', 'doorjamb', 'mouse', 'door', 'sink', 'napkin', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'measuring_cup', 'microwave', 'freezer', 'ceilingfan', 'milk', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'button', 'toaster', 'faucet', 'picture', 'wall', 'nightstand', 'oven', 'photoframe', 'clothes_gloves', 'coffe_maker', 'food_food', 'curtain', 'centerpiece', 'food_sugar', 'vacuum_cleaner', 'chessboard', 'coffee_cup', 'dishwasher', 'light']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[LOOKAT] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <milk>', '[GRAB] <milk>', '[PUTBACK] <milk> <coffe_maker>', '[FIND] <food_sugar>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[SWITCHON] <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_bread', 'tablelamp', 'crayon', 'walllamp', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'dvd_player', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'coffe_maker', 'television', 'curtain', 'after_shave', 'doorjamb', 'door', 'bathroom_cabinet', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "You walk to the living room and sit on the sofa. You grab the remote control, look at it and point at the television, you watch television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'food_ice_cream', 'tea', 'walllamp', 'drawing', 'bills', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'keyboard', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'iron', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'toy', 'television', 'tray', 'curtain', 'sink', 'bathroom_cabinet', 'table']\nTask:Read magazine\nresponse:\n{\n    \"thoughts\": Please sit on the couch and read a magazine,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <newspaper>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <newspaper>', '[LOOKAT] <newspaper>', '[READ] <newspaper>', '[READ] <newspaper>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'juice', 'toilet', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'phone', 'desk', 'stovefan', 'plate', 'faucet', 'window', 'mousepad', 'form', 'pillow', 'wall', 'wall_clock', 'oven', 'clothes_gloves', 'tray', 'diary', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'sink']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[TURNTO] <plate>", "[LOOKAT] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[WASH] <plate>", "[PUTBACK] <plate> <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'band-aids', 'freezer', 'walllamp', 'nail_polish', 'trashcan', 'dough', 'drawing', 'bills', 'cpuscreen', 'bathroom_counter', 'ceiling', 'phone', 'character', 'desk', 'purse', 'alcohol', 'clothes_jacket', 'toaster', 'mat', 'faucet', 'oil', 'couch', 'broom', 'oven', 'photoframe', 'tray', 'kitchen_counter', 'dresser', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find a cat in the kitchen and touch it,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'food_carrot', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'food_bacon', 'ceilinglamp', 'mat', 'tvstand', 'stereo', 'mousepad', 'couch', 'wall', 'powersocket', 'toy', 'photoframe', 'mouse', 'controller', 'door', 'table', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "turn on the computer and browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'toaster', 'mat', 'faucet', 'shower', 'pillow', 'wall', 'nightstand', 'bench', 'toy', 'coffe_maker', 'television', 'curtain', 'after_shave', 'doorjamb', 'towel_rack', 'kitchen_counter', 'chessboard', 'sink', 'home_office', 'bathroom_cabinet', 'table', 'instrument_violin']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'filing_cabinet', 'tablelamp', 'food_ice_cream', 'beer', 'toilet', 'drawing', 'hanger', 'floor', 'bookshelf', 'envelope', 'tvstand', 'shower', 'faucet', 'candle', 'dining_room', 'broom', 'wall', 'toy', 'cup', 'photoframe', 'clothes_gloves', 'light_bulb', 'wooden_spoon', 'curtain', 'doorjamb', 'chessboard', 'dishrack', 'door']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[TURNTO] <toilet>", "[LOOKAT] <toilet>", "[WALK] <toilet>", "[TURNTO] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'dog', 'freezer', 'orchid', 'drawing', 'wallshelf', 'ceiling', 'character', 'mail', 'clothes_jacket', 'computer', 'mat', 'faucet', 'window', 'closetdrawer', 'nightstand', 'wall_clock', 'oven', 'basket_for_clothes', 'coffe_maker', 'mouse', 'dresser', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'light', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Pick up my dirty clothes, load the washer, add detergent, turn on washer,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[RUN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_bread', 'beer', 'freezer', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_cake', 'alcohol', 'computer', 'clothes_jacket', 'ceilinglamp', 'tvstand', 'mousepad', 'microphone', 'towel', 'wall', 'powersocket', 'food_food', 'television', 'wine', 'doorjamb', 'mouse', 'pencil', 'door', 'light', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen and put groceries in fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'knifeblock', 'drawing', 'cpuscreen', 'clothes_underwear', 'knife', 'floor', 'ceiling', 'cleaning_solution', 'piano_bench', 'toaster', 'ceilinglamp', 'faucet', 'stereo', 'iron', 'wall', 'oven', 'cutting_board', 'tape', 'check', 'coin', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'bag', 'home_office', 'instrument_violin', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Washing hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[SCRUB] <hands_both>', '[SWITCHON] <faucet>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'stereo', 'tvstand', 'mousepad', 'couch', 'broom', 'wall', 'bed', 'check', 'toilet_paper', 'television', 'ice', 'kettle', 'mouse', 'door', 'sink', 'home_office', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go in my office, sit at my desk turn on the computer, then type on my keyboard and open my web browser.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'microwave', 'cat', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'computer', 'clothes_jacket', 'button', 'ceilinglamp', 'mat', 'window', 'couch', 'food_jam', 'wall', 'nightstand', 'cutting_board', 'cup', 'television', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'table', 'vase', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into bedrrom and to the sofa. Sit on the sofa and touch cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'food_butter', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'towel', 'wall', 'pasta', 'bench', 'cup', 'curtain', 'mouse', 'dresser', 'clothes_dress', 'door', 'table', 'tooth_paste']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "enter office, find computer, turn it on,", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TOUCH] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'shower', 'window', 'homework', 'mousepad', 'pillow', 'powersocket', 'photoframe', 'coffe_maker', 'doorjamb', 'kettle', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'clothes_dress', 'door', 'bag', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to bedroom and grab dirty pants and take to the washing machine in entrance hall. Put clothes in washing machine and then put in laundry detergent.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_pants>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'blow_dryer', 'detergent', 'cat', 'drawing', 'desk', 'thread', 'computer', 'button', 'table_cloth', 'ceilinglamp', 'tvstand', 'mousepad', 'broom', 'nightstand', 'oven', 'powersocket', 'box', 'newspaper', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'conditioner', 'bag', 'napkin', 'table', 'light', 'soap']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my cat, and I pet him", "actions": ["[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'remote_control', 'computer', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'clothes_socks', 'closetdrawer', 'couch', 'pillow', 'towel', 'wall', 'powersocket', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch and use the remote control to turn on the TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'toilet', 'jelly', 'walllamp', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'stereo', 'electric_shaver', 'microphone', 'wall', 'pasta', 'powersocket', 'oven', 'bed', 'television', 'doorjamb', 'mouse', 'hair', 'door', 'table', 'colander', 'instrument_violin']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "first i will go to living room and then take the remote and swtitch on television after that il change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[TOUCH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'novel', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'comb', 'door', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Find book, grab book, fin bookmark, sit down somewhere comfortable, open book, read book, close book after reading.,\n    \"actions\": ['[WALK] <desk>', '[FIND] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[FIND] <couch>', '[SIT] <couch>', '[GRAB] <novel>', '[READ] <novel>', '[PUTBACK] <bookmark> <novel>']\n}\n", "input": "\nScene_Objects:['freezer', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bedroom', 'keyboard', 'food_cake', 'computer', 'button', 'ceilinglamp', 'mousepad', 'wall', 'nightstand', 'powersocket', 'cup', 'food_food', 'check', 'doorjamb', 'lighter', 'mouse', 'shoes', 'dishrack', 'controller', 'door', 'light', 'scrabble']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "put away groceries in fridge. open the door, put on shelf, close the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_salt', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'mousepad', 'microphone', 'dining_room', 'nightstand', 'basket_for_clothes', 'bed', 'newspaper', 'coffe_maker', 'check', 'tray', 'curtain', 'doorjamb', 'mouse', 'headset', 'laundry_detergent', 'maindoor', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to laundry room. Find laundry basket. Grab laundry basket. Find washing machine. Walk to washing machine. Open washing machine. Find detergent. Grab detergent. Put in detergent. Put in clothes. Close washing machine. Turn on machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[PUTOBJBACK] <laundry_detergent>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'knifeblock', 'bills', 'bathroom_counter', 'ceiling', 'character', 'desk', 'bedroom', 'computer', 'bookmark', 'shoe-shine_kit', 'pillow', 'towel', 'pasta', 'oven', 'powersocket', 'toy', 'cup', 'photoframe', 'light_bulb', 'television', 'kitchen_counter', 'chessboard', 'comb', 'pajamas', 'food_steak', 'sink', 'instrument_violin', 'bag', 'tooth_paste']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the home office. Pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'walllamp', 'knifeblock', 'orchid', 'floor', 'phone', 'ceiling', 'bedroom', 'keyboard', 'food_bacon', 'stovefan', 'toaster', 'plate', 'faucet', 'window', 'homework', 'closetdrawer', 'wall', 'oven', 'ice', 'tray', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'hairdryer', 'sink', 'bathroom_cabinet', 'coffee']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[TURNTO] <plate>', '[LOOKAT] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WASH] <plate>', '[PUTBACK] <plate> <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'toaster', 'stovefan', 'plate', 'faucet', 'iron', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'food_food', 'wooden_spoon', 'tray', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'blow_dryer', 'milk', 'bills', 'orchid', 'floor', 'ceiling', 'character', 'food_cake', 'alcohol', 'stovefan', 'homework', 'closetdrawer', 'pillow', 'broom', 'wall', 'oven', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'kitchen_counter', 'shoes', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'iron', 'mousepad', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'video_game_controller', 'television', 'doorjamb', 'kettle', 'mouse', 'door', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to home office. Walk to computer. Switch on Computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'toilet', 'blow_dryer', 'bathroom', 'trashcan', 'orchid', 'dvd_player', 'keyboard', 'mail', 'computer', 'food_bacon', 'stovefan', 'ceilinglamp', 'picture', 'window', 'couch', 'clothes_pants', 'oven', 'photoframe', 'coffe_maker', 'food_food', 'television', 'food_onion', 'coffee_pot', 'sink', 'bathroom_cabinet', 'light', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": grab the vesseles and wash,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dining_room>', '[WASH] <bowl>', '[WALK] <detergent>', '[FIND] <detergent>', '[RINSE] <detergent>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'trashcan', 'floor', 'tea_bag', 'thread', 'toaster', 'ceilinglamp', 'plate', 'dish_soap', 'wall', 'bed', 'coffe_maker', 'food_food', 'check', 'oven_mitts', 'tray', 'doorjamb', 'pencil', 'kitchen_counter', 'hairbrush', 'cupboard', 'hair', 'controller', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'faucet', 'shower', 'tooth_paste', 'towel', 'wall', 'wall_clock', 'oven', 'tape', 'photoframe', 'clothes_gloves', 'light_bulb', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'light', 'vase', 'soap', 'home_office', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, open the faucet rinse the hands, wash the hands with soap and rinse again, then wipe with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'dough', 'bills', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'shampoo', 'bookshelf', 'computer', 'mat', 'window', 'food_noodles', 'bowl', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'cup', 'bed', 'toilet_paper', 'mouthwash', 'ice', 'doorjamb', 'vacuum_cleaner', 'dishrack', 'door', 'napkin', 'home_office']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Walk to bedroom. Walk to bed, lie in bed.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'band-aids', 'freezer', 'nail_polish', 'bathroom', 'cat', 'orchid', 'cpuscreen', 'phone', 'character', 'keyboard', 'mat', 'faucet', 'wall', 'oven', 'bed', 'keys', 'coffe_maker', 'mouthwash', 'food_onion', 'curtain', 'centerpiece', 'console', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tea', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'phone', 'keyboard', 'food_cake', 'toaster', 'stovefan', 'plate', 'mat', 'dish_soap', 'oil', 'pillow', 'wall', 'cup', 'photoframe', 'coffe_maker', 'curtain', 'centerpiece', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'mop', 'dog', 'jelly', 'drawing', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'character', 'computer', 'stovefan', 'faucet', 'oil', 'dining_room', 'wall', 'nightstand', 'oven', 'cutting_board', 'photoframe', 'check', 'tray', 'sink', 'napkin', 'bathroom_cabinet', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": go to the bathroom and wash your hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[SQUEEZE] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'sponge', 'tablelamp', 'food_ice_cream', 'crayon', 'measuring_cup', 'toilet', 'knifeblock', 'orchid', 'bathroom_counter', 'dvd_player', 'keyboard', 'computer', 'stovefan', 'mat', 'toaster', 'picture', 'window', 'microphone', 'pillow', 'bench', 'toy', 'novel', 'tray', 'curtain', 'chessboard', 'controller', 'conditioner', 'sink', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find cat. Walk to cat. Pull cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[PULL] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'laptop', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watching TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'band-aids', 'detergent', 'cat', 'knife', 'bathroom_counter', 'shampoo', 'floor', 'character', 'computer', 'bookmark', 'stovefan', 'mousepad', 'form', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'keys', 'coffe_maker', 'television', 'food_onion', 'curtain', 'centerpiece', 'dishrack', 'sink', 'light', 'clothes_skirt', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into living room, look at the phone when it ringing, grab the phone or pickup the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'walllamp', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'face', 'computer', 'bookmark', 'mat', 'faucet', 'mousepad', 'clothes_hat', 'towel', 'dining_room', 'wall', 'wall_clock', 'nightstand', 'face_soap', 'basket_for_clothes', 'newspaper', 'tray', 'rag', 'mouse', 'dishrack', 'sink', 'bathroom_cabinet']\nTask:Wash face\nresponse:\n{\n    \"thoughts\": Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <rag>', '[GRAB] <rag>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[WASH] <rag>', '[FIND] <face_soap>', '[GRAB] <face_soap>', '[POUR] <face_soap> <rag>', '[PUTOBJBACK] <face_soap>', '[FIND] <face>', '[SCRUB] <face>', '[PUTBACK] <rag> <sink>', '[RINSE] <face>', '[RINSE] <rag>', '[SWITCHOFF] <faucet>', '[GRAB] <rag>', '[PUTOBJBACK] <rag>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <face>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'ground_coffee', 'nail_polish', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'toaster', 'ceilinglamp', 'mat', 'faucet', 'stovefan', 'mousepad', 'oil', 'wall', 'bed', 'mouse', 'cupboard', 'door', 'hairdryer', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[PUSH] <mouse>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'filing_cabinet', 'beer', 'measuring_cup', 'chair', 'ground_coffee', 'knifeblock', 'longboard', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'mail', 'toaster', 'picture', 'dustpan', 'wall', 'bench', 'cutting_board', 'tape', 'cup', 'coffe_maker', 'coffee_filter', 'water', 'after_shave', 'dresser', 'maindoor', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[GRAB] <ground_coffee>', '[PUTOBJBACK] <ground_coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'walllamp', 'cat', 'drawing', 'orchid', 'knifeblock', 'knife', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'ceilinglamp', 'tvstand', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'photoframe', 'coffe_maker', 'check', 'curtain', 'lighter', 'dresser', 'napkin', 'home_office', 'table', 'coffee']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into bedrrom and to the sofa. Sit on the sofa and touch cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'crayon', 'walllamp', 'woman', 'knifeblock', 'hanger', 'ceiling', 'folder', 'character', 'man', 'mop_bucket', 'bookmark', 'shower', 'window', 'electric_shaver', 'pillow', 'dining_room', 'wall_clock', 'cutting_board', 'bed', 'television', 'wine', 'curtain', 'mouse', 'pencil', 'sink', 'home_office', 'light', 'dirt', 'shoe_rack']\nTask:Greet people\nresponse:\n{\n    \"thoughts\": Say HI to those that come in.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <woman>', '[FIND] <woman>', '[GREET] <woman>', '[FIND] <woman>', '[GREET] <woman>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'beer', 'chair', 'freezer', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mousepad', 'food_jam', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'food_onion', 'doorjamb', 'coffee_pot', 'mouse', 'kitchen_counter', 'dresser', 'door', 'table', 'light', 'dishwasher']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk to the bedroom, then find the computer and sit in the chair ,next switch on the computer and finally type in the keyboard to start to work on it", "actions": ["[WALK] <bedroom>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'nail_polish', 'knife', 'cpuscreen', 'floor', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'mousepad', 'closetdrawer', 'couch', 'dining_room', 'wall', 'basket_for_clothes', 'cup', 'check', 'television', 'curtain', 'mouse', 'chessboard', 'shoes', 'comb', 'cupboard', 'hairdryer', 'door', 'home_office']\nTask:Read book\nresponse:\n{\n    \"thoughts\": find a book to read. sit down. read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'ground_coffee', 'walllamp', 'trashcan', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'toaster', 'shoe-shine_kit', 'faucet', 'clothes_pants', 'dining_room', 'wall', 'photoframe', 'clothes_gloves', 'coffe_maker', 'coffee_filter', 'centerpiece', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'clothes_scarf', 'home_office', 'table', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open the coffee maker and put a filer, grab ground coffee and put it in the coffee maker, then close the coffee maker and switch it on", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'fork', 'beer', 'walllamp', 'bathroom', 'orchid', 'kitchen_cabinet', 'floor', 'phone', 'toaster', 'mat', 'plate', 'tvstand', 'window', 'dish_soap', 'couch', 'bowl', 'wall', 'oven', 'bench', 'coffe_maker', 'tray', 'rag', 'console', 'kitchen_counter', 'dishrack', 'hair', 'maindoor', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>', '[GRAB] <plate>', '[WIPE] <plate>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[PUTOBJBACK] <dishrack>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'dog', 'freezer', 'walllamp', 'trashcan', 'drawing', 'bills', 'floor', 'ceiling', 'bookshelf', 'computer', 'button', 'tvstand', 'stereo', 'wall', 'nightstand', 'bench', 'bed', 'photoframe', 'television', 'razor', 'food_rice', 'wine', 'tray', 'comb', 'sink', 'bathroom_cabinet', 'light', 'dirt']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into my bedroom and see my stereo. I decide to play music. I walk over to the stereo and turn it on. I sit down on the bed to listen to the music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'chair', 'freezer', 'detergent', 'bathroom_counter', 'desk', 'bedroom', 'keyboard', 'button', 'stovefan', 'ceilinglamp', 'stereo', 'shower', 'window', 'clothes_socks', 'dining_room', 'nightstand', 'wall', 'bench', 'oven', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'dresser', 'napkin', 'bathroom_cabinet', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room, look for radio, grab on button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'microwave', 'chair', 'ground_coffee', 'walllamp', 'detergent', 'trashcan', 'drawing', 'knife', 'floor', 'ceiling', 'thread', 'toaster', 'dustpan', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'chessboard', 'cupboard', 'spoon', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find coffee maker. Find coffee filter. Put it in the coffee maker. Find ground coffee. Put it in the coffee maker. Find Water. Put it in the coffe maker. Switch on the coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <water>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'bathroom', 'detergent', 'cat', 'food_butter', 'toothbrush', 'orchid', 'cpuscreen', 'desk', 'bedroom', 'purse', 'envelope', 'mail', 'mop_bucket', 'toaster', 'shower', 'faucet', 'oil', 'food_orange', 'powersocket', 'cutting_board', 'shelf', 'curtain', 'after_shave', 'diary', 'kitchen_counter', 'door', 'napkin', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": before going to bed or when wake up earily morning brush teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'toilet', 'chair', 'freezer', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'button', 'ceilinglamp', 'mat', 'foundation', 'mousepad', 'clothes_pants', 'wall', 'photoframe', 'television', 'mouse', 'kitchen_counter', 'headset', 'door', 'home_office', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to home office and work on computer.", "actions": ["[RUN] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TOUCH] <mouse>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'microwave', 'toilet', 'walllamp', 'drawing', 'bills', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'television', 'tray', 'curtain', 'towel_rack', 'pencil', 'kitchen_counter', 'bathtub', 'spoon', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'window', 'homework', 'couch', 'candle', 'pillow', 'wall', 'nightstand', 'powersocket', 'check', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room. Sit down on the sofa. Pick up the remote control. Turn on the television with the remote control. Put down the remote control. Watch the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'walllamp', 'nail_polish', 'detergent', 'dough', 'drawing', 'food_egg', 'tea_bag', 'bookshelf', 'remote_control', 'alcohol', 'button', 'mop_bucket', 'table_cloth', 'stovefan', 'tvstand', 'shoe-shine_kit', 'shower', 'stereo', 'deck_of_cards', 'mousepad', 'powersocket', 'check', 'rag', 'towel_rack', 'pencil', 'conditioner', 'door', 'bathroom_cabinet', 'soap']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'beer', 'walllamp', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'purse', 'food_bacon', 'piano_bench', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'shower', 'window', 'foundation', 'couch', 'pillow', 'wall', 'bench', 'check', 'wine', 'curtain', 'hairbrush', 'controller', 'door', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'dough', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'oil', 'wall', 'powersocket', 'shelf', 'toilet_paper', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'light', 'table', 'vase']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Keep my keyboard vacuumed and free of debris.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'scissors', 'chair', 'hanger', 'phone', 'keyboard', 'bookshelf', 'plate', 'faucet', 'window', 'mousepad', 'food_noodles', 'oil', 'food_peanut_butter', 'clothes_pants', 'dining_room', 'bench', 'oven', 'cutting_board', 'toy', 'basket_for_clothes', 'cup', 'photoframe', 'keys', 'food_food', 'laundry_detergent', 'bathroom_cabinet', 'instrument_violin', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'walllamp', 'nail_polish', 'detergent', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'couch', 'microphone', 'pillow', 'wall', 'pasta', 'cup', 'photoframe', 'bed', 'novel', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'kitchen_counter', 'home_office']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the living room and find the book you want to read. Take it and then go to the couch. Sit on it, open the book and start reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'crayon', 'drawing', 'needle', 'bills', 'knifeblock', 'floor', 'phone', 'ceiling', 'character', 'mat', 'plate', 'homework', 'pillow', 'bowl', 'bench', 'newspaper', 'novel', 'food_food', 'video_game_controller', 'television', 'oven_mitts', 'food_rice', 'coffee_pot', 'shoes', 'cd', 'home_office', 'table', 'teeth']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.", "actions": ["[WALK] <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'toaster', 'stovefan', 'plate', 'faucet', 'iron', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'food_food', 'wooden_spoon', 'tray', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'measuring_cup', 'microwave', 'walllamp', 'food_carrot', 'orchid', 'knifeblock', 'cpuscreen', 'drawing', 'floor', 'bedroom', 'cleaning_solution', 'computer', 'mop_bucket', 'table_cloth', 'stovefan', 'shower', 'iron', 'couch', 'bowl', 'dining_room', 'nightstand', 'cup', 'keys', 'curtain', 'doorjamb', 'comb', 'dresser', 'sink', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[SQUEEZE] <cleaning_solution>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'juice', 'chair', 'jelly', 'nail_polish', 'cat', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'iron', 'clothes_pants', 'food_kiwi', 'wall', 'photoframe', 'food_food', 'water', 'wooden_spoon', 'curtain', 'doorjamb', 'sink', 'bathroom_cabinet', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Wash the teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[TURNTO] <water>', '[POINTAT] <water>', '[FIND] <teeth>', '[WASH] <teeth>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'towel', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'television', 'curtain', 'doorjamb', 'shaving_cream', 'pencil', 'door', 'home_office', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room. Sit on the sofa. Grab the remote control and plug in TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[PLUGIN] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'freezer', 'ground_coffee', 'bathroom', 'ceilingfan', 'drawing', 'clothes_underwear', 'floor', 'food_cake', 'toaster', 'mousepad', 'dining_room', 'oven', 'tape', 'box', 'photoframe', 'coin', 'curtain', 'shaving_cream', 'coffee_pot', 'headset', 'dishrack', 'cupboard', 'sink', 'clothes_scarf', 'home_office', 'dishwasher', 'soap', 'board_game', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'food_salt', 'freezer', 'walllamp', 'bathroom', 'detergent', 'knifeblock', 'desk', 'bedroom', 'bookshelf', 'keyboard', 'clothes_jacket', 'toaster', 'stereo', 'electric_shaver', 'food_noodles', 'dining_room', 'broom', 'wall', 'oven', 'box', 'cup', 'curtain', 'shoes', 'vacuum_cleaner', 'sink', 'bag', 'dishwasher', 'creditcard']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to home office 1. Walk over to phone 1 on table 1.  Grab phone 1.  lift phone 1.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'tvstand', 'shower', 'closetdrawer', 'wall', 'wall_clock', 'oven', 'cup', 'standingmirror', 'check', 'television', 'centerpiece', 'console', 'mouse', 'vacuum_cleaner', 'shoes', 'comb', 'dresser', 'hairbrush', 'sink', 'bag', 'napkin', 'bathroom_cabinet', 'vase']\nTask:vacuum carpet\nresponse:\n{\n    \"thoughts\": I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'scissors', 'tea', 'freezer', 'blow_dryer', 'drawing', 'knifeblock', 'cpuscreen', 'clothes_underwear', 'character', 'desk', 'toaster', 'shower', 'picture', 'mousepad', 'candle', 'dining_room', 'bench', 'oven', 'basket_for_clothes', 'bed', 'check', 'television', 'toothbrush_holder', 'curtain', 'mouse', 'sink', 'light', 'tooth_paste', 'laser_pointer']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the living room and then I walk to the light switch in there and turn the lights on", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'filing_cabinet', 'scissors', 'chair', 'ceilingfan', 'dough', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'coffee_table', 'television', 'coin', 'wooden_spoon', 'curtain', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['band-aids', 'walllamp', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'mail', 'bookmark', 'plate', 'faucet', 'picture', 'dish_soap', 'dustpan', 'wall', 'oven', 'bench', 'bed', 'food_food', 'curtain', 'doorjamb', 'kettle', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'cupboard', 'door', 'sink', 'dirt']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "i go to the kitchen, i find the faucet. I turn the faucet on, i grab the plate to wash , i then get the soap. I wash the dishes by hand", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'envelope', 'computer', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'photoframe', 'check', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to the home office to work on my project. I turn on the light and walk to my desk. I turn on the computer. Then I look through the documents. After I'm done, I turn off the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <check>', '[TURNTO] <check>', '[LOOKAT] <check>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['toilet', 'blow_dryer', 'bathroom', 'walllamp', 'orchid', 'cpuscreen', 'dvd_player', 'character', 'desk', 'stamp', 'envelope', 'stovefan', 'ceilinglamp', 'shoe-shine_kit', 'shower', 'homework', 'food_orange', 'dining_room', 'oven', 'bench', 'box', 'video_game_controller', 'television', 'food_donut', 'curtain', 'doorjamb', 'pajamas', 'napkin', 'light', 'clothes_shirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'chair', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'mat', 'mousepad', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'food_food', 'coffe_maker', 'doorjamb', 'pencil', 'kitchen_counter', 'cd', 'shoes', 'hair', 'clothes_dress', 'table']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'toilet', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'phone', 'character', 'piano_bench', 'bookmark', 'toaster', 'mat', 'shower', 'dish_soap', 'couch', 'wall', 'photoframe', 'novel', 'coffe_maker', 'check', 'cards', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'sink', 'home_office', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk to kitchen, find the dishwasher, apply dish soap into vessels or dishwasher, wash the dishes with a dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[WASH] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'keyboard', 'computer', 'toaster', 'tvstand', 'floor_lamp', 'window', 'closetdrawer', 'couch', 'towel', 'wall', 'wall_clock', 'food_food', 'doorjamb', 'mouse', 'headset', 'shoes', 'dresser', 'controller', 'bathtub', 'sink', 'home_office', 'light']\nTask:Turn night light on\nresponse:\n{\n    \"thoughts\": I walk to the dresser and in the middle Io to small battery operated lamp and switch it on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <floor_lamp>', '[TURNTO] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'chair', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'bookshelf', 'mat', 'laptop', 'faucet', 'wall', 'box', 'photoframe', 'coffe_maker', 'colander', 'curtain', 'pencil', 'headset', 'dishrack', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'clothes_shirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk in to the bathroom. I find the soap. I find the faucet. I turn on the faucet. I wash my hands. I put the soap back. I rinse my hands. I turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'filing_cabinet', 'crayon', 'walllamp', 'bathroom', 'toothbrush', 'floor', 'food_cake', 'toaster', 'plate', 'shower', 'faucet', 'electric_shaver', 'dish_soap', 'broom', 'wall', 'oven', 'bench', 'photoframe', 'light_bulb', 'novel', 'coffe_maker', 'television', 'tray', 'towel_rack', 'kettle', 'dishrack', 'table', 'dishwasher', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['blender', 'walllamp', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'thread', 'bookshelf', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'clothes_socks', 'wall', 'pasta', 'powersocket', 'cutting_board', 'cup', 'clothes_gloves', 'novel', 'television', 'rag', 'doorjamb', 'shoes', 'door', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "turn on the tv", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'ground_coffee', 'nail_polish', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'toaster', 'ceilinglamp', 'mat', 'faucet', 'stovefan', 'mousepad', 'oil', 'wall', 'bed', 'mouse', 'cupboard', 'door', 'hairdryer', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['crayon', 'toilet', 'chair', 'bathroom', 'detergent', 'dough', 'knife', 'drawing', 'floor', 'ceiling', 'tea_bag', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'mousepad', 'clothes_socks', 'pillow', 'wall', 'cup', 'doorjamb', 'console', 'shoes', 'dresser', 'door', 'hairdryer', 'napkin', 'light', 'soap', 'laser_pointer']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch. Turn on lights.", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_bread', 'bathroom', 'trashcan', 'food_carrot', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'phone', 'folder', 'bedroom', 'desk', 'keyboard', 'toaster', 'bench', 'tray', 'doorjamb', 'diary', 'console', 'mouse', 'towel_rack', 'cd', 'dresser', 'cupboard', 'controller', 'conditioner', 'table', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to table, grab water glass and drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>', '[PUTBACK] <water_glass> <table>']\n}\n", "input": "\nScene_Objects:['spectacles', 'blender', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'computer', 'ceilinglamp', 'homework', 'mousepad', 'dining_room', 'food_kiwi', 'wall', 'powersocket', 'bench', 'oven', 'bed', 'food_food', 'doorjamb', 'mouse', 'controller', 'door', 'bathroom_cabinet', 'light', 'soap', 'clothes_shirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen, open fridge, get groceries, put groceries in fridge, close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'trashcan', 'drawing', 'floor', 'dvd_player', 'desk', 'envelope', 'ceilinglamp', 'tvstand', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'cup', 'check', 'television', 'coin', 'colander', 'after_shave', 'cd', 'cupboard', 'table', 'vase', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'crayon', 'chair', 'walllamp', 'detergent', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'pillow', 'wall', 'powersocket', 'newspaper', 'television', 'tray', 'doorjamb', 'mouse', 'pencil', 'door', 'light', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I walk into my office and go to the computer desk. I sit at my desk and turn the computer on. Then I write an email.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'detergent', 'cat', 'milk', 'phone', 'ceiling', 'desk', 'mop_bucket', 'brush', 'ceilinglamp', 'tvstand', 'faucet', 'toaster', 'mat', 'closetdrawer', 'microphone', 'nightstand', 'wall_clock', 'toy', 'cup', 'food_food', 'after_shave', 'lighter', 'towel_rack', 'mouse', 'pajamas', 'controller', 'laundry_detergent', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Go to the living room where the cat sleeps. Find cat. Proceed to petting cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'water_glass', 'pot', 'measuring_cup', 'microwave', 'walllamp', 'blow_dryer', 'trashcan', 'knife', 'knifeblock', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'ceilinglamp', 'dish_soap', 'wall', 'oven', 'cup', 'bed', 'coffe_maker', 'tray', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'coffee_cup', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and put plates, cutlery and glasses into the dishwasher, put the soap, close the machine and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'tablelamp', 'juice', 'food_salt', 'microwave', 'chair', 'freezer', 'detergent', 'trashcan', 'toothbrush', 'floor', 'ceiling', 'bookshelf', 'toaster', 'shower', 'stereo', 'mousepad', 'clothes_pants', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'food_food', 'check', 'food_donut', 'kitchen_counter', 'cupboard', 'door', 'home_office']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Go kitchen and make something for eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'juice', 'milk', 'hanger', 'kitchen_cabinet', 'bathroom_counter', 'button', 'food_bacon', 'stovefan', 'mat', 'shoe-shine_kit', 'tvstand', 'electric_shaver', 'food_peanut_butter', 'form', 'dining_room', 'cutting_board', 'cup', 'photoframe', 'coffe_maker', 'television', 'water', 'wooden_spoon', 'kettle', 'kitchen_counter', 'chessboard', 'dresser', 'sink', 'spoon']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to kitchen. Open the the cabinet. Locate the water glass. Pour water into glass. Drag the bottle close to mouth. Drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[CLOSE] <kitchen_cabinet>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'button', 'clothes_jacket', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'powersocket', 'bed', 'check', 'food_food', 'food_onion', 'dresser', 'light']\nTask:Put on glasses\nresponse:\n{\n    \"thoughts\": Put on glasses.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[FIND] <spectacles>', '[GRAB] <spectacles>', '[PUTON] <spectacles>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'ground_coffee', 'drawing', 'orchid', 'needle', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'wall', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'hair', 'clothes_dress', 'door', 'table', 'clothes_skirt', 'scrabble']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'photoframe', 'novel', 'television', 'mouse', 'kitchen_counter', 'door', 'sink', 'clothes_skirt', 'clothes_shirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to home office, find the book. Sit in the chair and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_bread', 'walllamp', 'toothbrush', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'toaster', 'ceilinglamp', 'window', 'homework', 'food_noodles', 'couch', 'candle', 'pillow', 'dining_room', 'wall', 'powersocket', 'toy', 'food_food', 'toilet_paper', 'tray', 'curtain', 'food_onion', 'shoes', 'door', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'headset', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <chair>', '[LOOKAT] <chair>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'mousepad', 'pillow', 'wall', 'nightstand', 'photoframe', 'razor', 'television', 'food_rice', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathroom_cabinet', 'table', 'dirt']\nTask:Work\nresponse:\n", "output": {"thoughts": "I go to my office. I switch the computer on and sit on the office chair. I look at the computer screen to see what needs to be done. I use my mouse to navigate to the relevant directory, then release it. I type code with my keyboard. When I'm done, I look at the computer again to evaluate what I've done. Then I switch off the computer using the mouse and put back the mouse.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[GRAB] <mouse>", "[SWITCHOFF] <computer>", "[PUTOBJBACK] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'food_salt', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'computer', 'mat', 'faucet', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'toothbrush_holder', 'wine', 'doorjamb', 'shaving_cream', 'kitchen_counter', 'door', 'sink', 'clothes_scarf', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into living room. Find wall. Find light button. Switch on light button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'chair', 'freezer', 'cat', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'stereo', 'faucet', 'picture', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'cup', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'home_office', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to the bedroom and sit at my desk. I turn my radio on and listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'chair', 'walllamp', 'cat', 'wallshelf', 'floor', 'character', 'desk', 'tea_bag', 'toaster', 'plate', 'couch', 'dining_room', 'box', 'newspaper', 'bed', 'toilet_paper', 'curtain', 'coffee_pot', 'mouse', 'kettle', 'cupboard', 'controller', 'clothes_dress', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find Cat in living room.  Walk to cat.  Pet cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'microwave', 'chair', 'walllamp', 'trashcan', 'food_butter', 'orchid', 'floor', 'dvd_player', 'desk', 'toaster', 'tvstand', 'oil', 'wall', 'bench', 'basket_for_clothes', 'photoframe', 'coffe_maker', 'razor', 'wooden_spoon', 'curtain', 'coffee_pot', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'spoon', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dishrack>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <dishrack> <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'beer', 'chair', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'dvd_player', 'stamp', 'bookshelf', 'toaster', 'ceilinglamp', 'faucet', 'homework', 'floor_lamp', 'window', 'food_orange', 'dining_room', 'nightstand', 'bench', 'food_cheese', 'television', 'doorjamb', 'lighter', 'cupboard', 'home_office', 'dishwasher', 'laser_pointer']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Walk to the lamp.  Find the switch.  Click the switch off.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[SWITCHOFF] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'toilet', 'trashcan', 'food_carrot', 'food_egg', 'ceiling', 'folder', 'stamp', 'bookshelf', 'toaster', 'laptop', 'shower', 'food_noodles', 'couch', 'form', 'nightstand', 'powersocket', 'bench', 'toy', 'bed', 'keys', 'novel', 'doorjamb', 'hairbrush', 'cupboard', 'conditioner', 'spoon', 'bathroom_cabinet', 'table', 'board_game']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the living room, walk over to phone, grab the phone, put it to your ear .", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'microwave', 'toilet', 'jelly', 'drawing', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'ceilinglamp', 'deck_of_cards', 'stereo', 'shower', 'window', 'couch', 'candle', 'microphone', 'wall', 'powersocket', 'box', 'photoframe', 'television', 'tray', 'towel_rack', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'home_office']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'trashcan', 'food_carrot', 'knifeblock', 'drawing', 'cpuscreen', 'clothes_underwear', 'floor', 'bedroom', 'alcohol', 'laptop', 'tvstand', 'stereo', 'faucet', 'mat', 'couch', 'clothes_hat', 'nightstand', 'bench', 'oven', 'toy', 'coffe_maker', 'television', 'coin', 'kitchen_counter', 'cupboard', 'home_office', 'bathroom_cabinet', 'dishwasher', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "go to radio. run on radio.", "actions": ["[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'measuring_cup', 'walllamp', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'food_noodles', 'couch', 'broom', 'wall', 'newspaper', 'light_bulb', 'novel', 'television', 'curtain', 'doorjamb', 'lighter', 'cupboard', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['beer', 'bathroom', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'button', 'brush', 'mat', 'toaster', 'faucet', 'picture', 'iron', 'pillow', 'broom', 'wall', 'oven', 'keys', 'novel', 'coin', 'tray', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'soap', 'clothes_shirt', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[PUSH] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'face', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'wall_clock', 'face_soap', 'food_food', 'coffe_maker', 'tray', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'shoe_rack']\nTask:Wash face\nresponse:\n{\n    \"thoughts\": Run warm water. Cleanse my face with soap. Rinse my face with water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <face_soap>', '[GRAB] <face_soap>', '[FIND] <face>', '[SCRUB] <face>', '[PUTOBJBACK] <face_soap>', '[WASH] <face>', '[RINSE] <face>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'ground_coffee', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'bedroom', 'ceilinglamp', 'tvstand', 'stereo', 'food_orange', 'cup', 'bed', 'keys', 'novel', 'food_food', 'coffe_maker', 'mouthwash', 'coin', 'check', 'diary', 'coffee_pot', 'mouse', 'dresser', 'hair', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to my bedroom, and then I switch on the radio", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'toilet', 'nail_polish', 'detergent', 'cat', 'trashcan', 'milk', 'knifeblock', 'phone', 'ceiling', 'desk', 'keyboard', 'mousepad', 'couch', 'pillow', 'bench', 'oven', 'cup', 'bed', 'light_bulb', 'food_food', 'lighter', 'kitchen_counter', 'hairbrush', 'dresser', 'laundry_detergent', 'table', 'clothes_skirt', 'tooth_paste', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Petting a cat may sound simple.Letting the Cat Come to You.Extend a hand or finger and allow the cat a chance to touch her nose to you. When a cat bumps her head into your hand, it's a signal she wants attention.Focusing on Areas with Scent Glands. Feed with food.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <milk>', '[FIND] <milk>', '[GRAB] <milk>', '[POUR] <milk> <cup>', '[TURNTO] <cat>', '[POINTAT] <cat>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sponge', 'food_bread', 'blender', 'chair', 'walllamp', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'tvstand', 'window', 'mousepad', 'clothes_socks', 'wall', 'oven', 'cup', 'television', 'mouse', 'cupboard', 'door', 'home_office', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "go to the office, turn computer on, open an browser or any application to open any website.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'band-aids', 'freezer', 'bathroom', 'detergent', 'food_butter', 'drawing', 'wallshelf', 'bookshelf', 'computer', 'tvstand', 'shower', 'homework', 'clothes_pants', 'bowl', 'nightstand', 'wall', 'oven', 'cup', 'photoframe', 'food_food', 'toilet_paper', 'coffe_maker', 'wine', 'tray', 'kitchen_counter', 'cupboard', 'door', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find the cat.  Pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'toilet', 'freezer', 'cpuscreen', 'ceiling', 'desk', 'computer', 'button', 'piano_bench', 'mop_bucket', 'mat', 'stereo', 'food_noodles', 'mousepad', 'wall', 'bench', 'bed', 'novel', 'oven_mitts', 'curtain', 'doorjamb', 'centerpiece', 'kettle', 'cupboard', 'table', 'dishwasher', 'instrument_violin', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "grab the vesseles and wash", "actions": ["[WALK] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dining_room>", "[WASH] <bowl>", "[WALK] <detergent>", "[FIND] <detergent>", "[RINSE] <detergent>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'mop', 'microwave', 'walllamp', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'iron', 'window', 'laptop', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'check', 'tray', 'rag', 'towel_rack', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Run hot water into tub of dirty dishes in the sink.  Add liquid soap.  Scrub dishes with wash rag. Rinse and place on drying rack.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <rag>', '[GRAB] <rag>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'tea', 'microwave', 'chair', 'walllamp', 'blow_dryer', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'plate', 'mousepad', 'food_noodles', 'couch', 'closetdrawer', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'razor', 'television', 'shaving_cream', 'kitchen_counter', 'bathroom_cabinet', 'dishwasher', 'soap', 'teeth']\nTask:Text friends while sitting on couch\nresponse:\n", "output": {"thoughts": "Text while sitting", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[TOUCH] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'knife', 'drawing', 'wallshelf', 'orchid', 'hanger', 'cpuscreen', 'floor', 'character', 'keyboard', 'purse', 'homework', 'mousepad', 'wall', 'pasta', 'bench', 'keys', 'television', 'wooden_spoon', 'towel_rack', 'kettle', 'kitchen_counter', 'hair', 'conditioner', 'home_office', 'table']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'microwave', 'toilet', 'dog', 'detergent', 'orchid', 'wallshelf', 'food_egg', 'cpuscreen', 'phone', 'stamp', 'bedroom', 'bookshelf', 'bookmark', 'shower', 'mousepad', 'food_jam', 'towel', 'pasta', 'bed', 'novel', 'food_food', 'curtain', 'centerpiece', 'shaving_cream', 'dresser', 'cupboard', 'table', 'creditcard']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'crayon', 'tea', 'toilet', 'bathroom', 'orchid', 'hanger', 'floor', 'character', 'bedroom', 'tea_bag', 'keyboard', 'purse', 'toaster', 'mat', 'deck_of_cards', 'window', 'food_orange', 'pillow', 'photoframe', 'food_food', 'coffe_maker', 'food_donut', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'sink']\nTask:Open door\nresponse:\n{\n    \"thoughts\": Open the door when leaving or coming in.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'mat', 'plate', 'faucet', 'window', 'dish_soap', 'food_noodles', 'closetdrawer', 'couch', 'towel', 'dining_room', 'wall', 'wine', 'rag', 'shaving_cream', 'kitchen_counter', 'chessboard', 'dresser', 'cupboard', 'conditioner', 'sink', 'table', 'dishwasher', 'soap']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Run warm water in the sink. Pour dish washing liquid on the dishes. Pick up dishes and wash them individually with a sponge. Rinse suds off under running water. Dry dishes with towel.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[WASH] <plate>", "[RINSE] <plate>", "[RINSE] <sponge>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <sponge>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <plate>", "[PUTOBJBACK] <towel>", "[WALK] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[PUTIN] <plate> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'ground_coffee', 'walllamp', 'bathroom', 'trashcan', 'floor', 'ceiling', 'phone', 'computer', 'toaster', 'stovefan', 'homework', 'wall', 'novel', 'coffe_maker', 'television', 'coffee_filter', 'water', 'tray', 'after_shave', 'diary', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'food_steak', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'freezer', 'walllamp', 'detergent', 'dough', 'floor', 'phone', 'ceiling', 'tea_bag', 'bookshelf', 'keyboard', 'toaster', 'laptop', 'tvstand', 'faucet', 'mousepad', 'dining_room', 'wall', 'box', 'cup', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'dishwasher']\nTask:Make popcorn\nresponse:\n", "output": {"thoughts": "Lay popcorn bag flat inside microwave and press popcorn button to start.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <microwave>", "[TURNTO] <microwave>", "[OPEN] <microwave>", "[PUTIN] <food_food> <microwave>", "[CLOSE] <microwave>", "[SWITCHON] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'envelope', 'computer', 'mop_bucket', 'stovefan', 'mat', 'tvstand', 'toaster', 'foundation', 'mousepad', 'pillow', 'wall', 'standingmirror', 'photoframe', 'toilet_paper', 'oven_mitts', 'tray', 'mouse', 'sink']\nTask:Read book\nresponse:\n{\n    \"thoughts\": go to living room,find the text book on the self ,sat on the chair ,open and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'beer', 'toilet', 'chair', 'band-aids', 'walllamp', 'bathroom', 'detergent', 'bills', 'knife', 'knifeblock', 'hanger', 'bedroom', 'button', 'plate', 'tvstand', 'iron', 'window', 'food_noodles', 'candle', 'nightstand', 'child', 'check', 'food_food', 'toothbrush_holder', 'razor', 'food_donut', 'cupboard', 'home_office', 'bathroom_cabinet']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Say hi", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'detergent', 'trashcan', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'closetdrawer', 'wall', 'oven', 'cup', 'novel', 'water', 'drinking_glass', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'door']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to the kitchen,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <water>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'keyboard', 'mail', 'mat', 'faucet', 'wall', 'nightstand', 'powersocket', 'oven', 'cup', 'photoframe', 'bed', 'novel', 'curtain', 'doorjamb', 'diary', 'cupboard', 'door', 'sink', 'instrument_guitar']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk over to the bookshelf. Pick storybook off the shelf. Sit on bed. Open storybook and read out loud.", "actions": ["[WALK] <bedroom>", "[WALK] <bookshelf>", "[FIND] <novel>", "[GRAB] <novel>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[WALK] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'crayon', 'microwave', 'food_butter', 'drawing', 'knifeblock', 'floor', 'bedroom', 'envelope', 'stovefan', 'ceilinglamp', 'mat', 'balanceball', 'window', 'couch', 'clothes_hat', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'television', 'after_shave', 'cards', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'hairdryer', 'spoon', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Walk to the sofa and sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'knifeblock', 'knife', 'orchid', 'floor', 'ceiling', 'phone', 'keyboard', 'mail', 'stovefan', 'ceilinglamp', 'deck_of_cards', 'stereo', 'pillow', 'dining_room', 'wall', 'cup', 'bed', 'novel', 'coffe_maker', 'doorjamb', 'kitchen_counter', 'cupboard', 'controller', 'laundry_detergent', 'hairdryer', 'sink', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'newspaper', 'television', 'food_donut', 'food_rice', 'doorjamb', 'mouse', 'cd', 'dresser']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'fork', 'scissors', 'food_ice_cream', 'microwave', 'walllamp', 'knifeblock', 'knife', 'wallshelf', 'floor', 'ceiling', 'phone', 'alcohol', 'clothes_jacket', 'toaster', 'stovefan', 'faucet', 'couch', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'rag', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'colander']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Take out a rag. Turn on faucet and get water on the rag. Clean counter off.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <rag> <sink>", "[SWITCHOFF] <faucet>", "[GRAB] <rag>", "[SQUEEZE] <rag>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_salt', 'toilet', 'chair', 'bathroom', 'trashcan', 'drawing', 'floor', 'character', 'desk', 'purse', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'foundation', 'couch', 'pillow', 'nightstand', 'powersocket', 'novel', 'tray', 'mouse', 'bathroom_cabinet', 'soap', 'teeth']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'juice', 'microwave', 'knifeblock', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'piano_bench', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'stovefan', 'tvstand', 'dish_soap', 'wall', 'pasta', 'coffee_filter', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'sink', 'scrabble']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <sink>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <sink>", "[SWITCHOFF] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[GRAB] <plate>", "[SCRUB] <plate>", "[PUTOBJBACK] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[PUTOBJBACK] <plate>", "[SWITCHON] <faucet>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[GRAB] <plate>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'laptop', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'tape', 'cup', 'photoframe', 'food_food', 'mouse', 'vacuum_cleaner', 'chessboard', 'napkin', 'table', 'soap']\nTask:Work\nresponse:\n{\n    \"thoughts\": first ill go to my home office adn ill switch on my laptop and with the use of mouse and keyboard il run my computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[RUN] <laptop>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'walllamp', 'cat', 'drawing', 'needle', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'mat', 'ceilinglamp', 'faucet', 'stereo', 'towel', 'pillow', 'wall', 'tape', 'bed', 'toilet_paper', 'video_game_controller', 'television', 'food_food', 'tray', 'diary', 'console', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Cleaning hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'cd_player', 'chair', 'knifeblock', 'music_stand', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'envelope', 'food_bacon', 'mop_bucket', 'stovefan', 'ceilinglamp', 'shower', 'nightstand', 'bed', 'photoframe', 'coffe_maker', 'tray', 'curtain', 'lighter', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to the living and turn to the music stand. Switch on the CD player to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <music_stand>', '[FIND] <music_stand>', '[TURNTO] <music_stand>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'walllamp', 'dough', 'knife', 'needle', 'food_egg', 'knifeblock', 'phone', 'ceiling', 'desk', 'bookshelf', 'stovefan', 'stereo', 'oil', 'food_orange', 'dining_room', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'wooden_spoon', 'curtain', 'doorjamb', 'kitchen_counter', 'sink', 'napkin', 'bag', 'light', 'dishwasher']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'dog', 'chair', 'bathroom', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'mat', 'ceilinglamp', 'shower', 'candle', 'wall', 'bed', 'clothes_gloves', 'food_food', 'water', 'food_rice', 'wooden_spoon', 'curtain', 'console', 'dishrack', 'pajamas', 'light', 'clothes_skirt', 'soap', 'dirt']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'microwave', 'toilet', 'jelly', 'freezer', 'walllamp', 'detergent', 'knifeblock', 'drawing', 'kitchen_cabinet', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'oil', 'wall', 'nightstand', 'oven', 'powersocket', 'photoframe', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Cooking a meal", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <pot> <sink>", "[GRAB] <pot>", "[SWITCHOFF] <faucet>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[CLOSE] <freezer>", "[WALK] <oven>", "[PUTBACK] <food_food> <pot>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'bathroom', 'drawing', 'needle', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stereo', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'toothbrush_holder', 'wooden_spoon', 'doorjamb', 'shoes', 'dresser', 'door', 'light', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on the light,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'mop', 'walllamp', 'detergent', 'food_carrot', 'toothbrush', 'knifeblock', 'orchid', 'clothes_underwear', 'hanger', 'phone', 'stovefan', 'shower', 'homework', 'couch', 'food_jam', 'microphone', 'bowl', 'photoframe', 'check', 'curtain', 'console', 'mouse', 'pencil', 'dresser', 'sink', 'soap', 'creditcard']\nTask:Pet dog\nresponse:\n", "output": {"thoughts": "Pet my dog in hall", "actions": ["[WALK] <home_office>", "[WALK] <dog>", "[TURNTO] <dog>", "[LOOKAT] <dog>", "[TOUCH] <dog>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'freezer', 'blow_dryer', 'bathroom', 'cpuscreen', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'food_bacon', 'tvstand', 'clothes_socks', 'oil', 'towel', 'nightstand', 'oven', 'bed', 'check', 'coffee_filter', 'curtain', 'clothes_dress', 'door', 'sink', 'home_office', 'table', 'light', 'colander', 'laser_pointer']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Pour water into a glass and drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'freezer', 'knife', 'floor', 'ceiling', 'folder', 'stamp', 'brush', 'mat', 'ceilinglamp', 'shower', 'faucet', 'toaster', 'pillow', 'bowl', 'food_kiwi', 'wall', 'powersocket', 'toilet_paper', 'mouthwash', 'doorjamb', 'door', 'home_office', 'light', 'table']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'detergent', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'purse', 'toaster', 'mat', 'faucet', 'iron', 'mousepad', 'wall', 'powersocket', 'cup', 'coffe_maker', 'water', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'clothes_scarf', 'colander']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'measuring_cup', 'microwave', 'blow_dryer', 'bathroom', 'woman', 'knifeblock', 'clothes_underwear', 'phone', 'dvd_player', 'man', 'mail', 'toaster', 'iron', 'faucet', 'food_orange', 'child', 'powersocket', 'cup', 'food_cheese', 'check', 'tray', 'curtain', 'door', 'sink', 'home_office', 'dishwasher', 'clothes_skirt', 'tooth_paste']\nTask:Say goodbye to guests leaving\nresponse:\n", "output": {"thoughts": "Tell them bye", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[FIND] <man>", "[TOUCH] <man>", "[FIND] <woman>", "[TOUCH] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'plate', 'faucet', 'couch', 'pillow', 'dining_room', 'broom', 'wall', 'powersocket', 'box', 'photoframe', 'novel', 'food_food', 'television', 'food_donut', 'curtain', 'doorjamb', 'bathroom_cabinet']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on couch. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <couch>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'bathroom', 'drawing', 'knife', 'clothes_underwear', 'floor', 'mail', 'food_bacon', 'piano_bench', 'stovefan', 'ceilinglamp', 'plate', 'deck_of_cards', 'toaster', 'oil', 'wall', 'wall_clock', 'oven', 'cup', 'novel', 'food_food', 'television', 'oven_mitts', 'cards', 'cupboard', 'conditioner', 'table']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "Sit down at the dinning table with my family and eat dinner.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[GRAB] <food_food>", "[FIND] <plate>", "[PUTBACK] <food_food> <plate>", "[GRAB] <food_food>", "[PUTOBJBACK] <food_food>", "[FIND] <fork>", "[GRAB] <fork>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'keyboard', 'computer', 'clothes_jacket', 'ceilinglamp', 'faucet', 'shower', 'window', 'towel', 'wall', 'newspaper', 'curtain', 'towel_rack', 'pencil', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'juice', 'microwave', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'envelope', 'mop_bucket', 'toaster', 'mat', 'faucet', 'dish_soap', 'couch', 'wall', 'cup', 'photoframe', 'toilet_paper', 'video_game_controller', 'wooden_spoon', 'tray', 'console', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'light']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sink>", "[PUTOBJBACK] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <cup>", "[GRAB] <cup>", "[SCRUB] <cup>", "[PUTOBJBACK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[SCRUB] <cup>", "[PUTOBJBACK] <cup>", "[PUTBACK] <sponge> <sink>", "[SWITCHON] <faucet>", "[RINSE] <cup>", "[GRAB] <cup>", "[FIND] <dishrack>", "[PUTBACK] <cup> <dishrack>", "[RINSE] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishrack>", "[RINSE] <sponge>", "[SWITCHOFF] <faucet>", "[GRAB] <sponge>", "[PUTOBJBACK] <sponge>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'trashcan', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'purse', 'mail', 'toaster', 'tvstand', 'faucet', 'towel', 'wall', 'nightstand', 'powersocket', 'food_food', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Turn faucet clockwise to turn on water.  Place hands under stream of water, pick up soap, rub hands with soap.  Put soap back down.  Rub hands together with soap on them.  Rinse hands under water from faucet.  Turn faucet counterclockwise to turn water off.  Dry hands by rubbing them on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[WASH] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'jelly', 'ground_coffee', 'knifeblock', 'drawing', 'needle', 'food_egg', 'hanger', 'clothes_underwear', 'desk', 'purse', 'alcohol', 'mat', 'stereo', 'faucet', 'food_peanut_butter', 'wall', 'powersocket', 'photoframe', 'novel', 'coffe_maker', 'curtain', 'mouse', 'vacuum_cleaner', 'shoes', 'bathroom_cabinet', 'clothes_shirt', 'electrical_outlet']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug vacuum into the wall. Turn on the vacuum. Move vacuum around the floor to clean it. Unplug vacuum.", "actions": ["[WALK] <home_office>", "[WALK] <vacuum_cleaner>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[FIND] <electrical_outlet>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[PUTOBJBACK] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'filing_cabinet', 'pot', 'tablelamp', 'measuring_cup', 'freezer', 'walllamp', 'ceilingfan', 'knifeblock', 'drawing', 'longboard', 'floor', 'ceiling', 'keyboard', 'button', 'clothes_jacket', 'tvstand', 'couch', 'form', 'dining_room', 'bowl', 'nightstand', 'light_bulb', 'check', 'curtain', 'mouse', 'hair', 'cupboard', 'hairdryer', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'orchid', 'knifeblock', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'television', 'curtain', 'shaving_cream', 'dresser', 'home_office', 'table', 'bathroom_cabinet', 'shoe_rack']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch favorite shows, netflix, hulu", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTBACK] <remote_control> <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mop_bucket', 'mat', 'laptop', 'window', 'homework', 'mousepad', 'food_orange', 'dining_room', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'bed', 'tray', 'curtain', 'mouse', 'pencil', 'light']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Bring laptop to bed and watch a movie,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[WALK] <bed>', '[SIT] <bed>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'character', 'video_game_console', 'bedroom', 'computer', 'button', 'table_cloth', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'box', 'video_game_controller', 'television', 'curtain', 'pencil', 'dresser', 'door', 'table']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'band-aids', 'dough', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'dvd_player', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'powersocket', 'cup', 'bed', 'photoframe', 'newspaper', 'water', 'food_donut', 'doorjamb', 'mouse', 'door', 'light', 'table', 'shoe_rack']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'food_salt', 'toilet', 'detergent', 'cpuscreen', 'food_egg', 'bathroom_counter', 'floor', 'folder', 'desk', 'envelope', 'clothes_jacket', 'mat', 'plate', 'stereo', 'deck_of_cards', 'shower', 'couch', 'television', 'oven_mitts', 'curtain', 'shaving_cream', 'chessboard', 'cupboard', 'conditioner', 'light', 'table', 'soap', 'clothes_shirt']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Get a broom or duster with long handle. Dust blades of ceiling fan.", "actions": ["[WALK] <bedroom>", "[WALK] <duster>", "[FIND] <duster>", "[GRAB] <duster>", "[PUTOBJBACK] <duster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'toilet', 'walllamp', 'detergent', 'food_carrot', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'bedroom', 'keyboard', 'bookshelf', 'mat', 'mousepad', 'dining_room', 'wall', 'nightstand', 'bench', 'basket_for_clothes', 'bed', 'photoframe', 'coffe_maker', 'food_food', 'wooden_spoon', 'mouse', 'dresser', 'sink', 'dishwasher']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go t the bedroom and go to sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'tea', 'toilet', 'chair', 'freezer', 'drawing', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'dvd_player', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'food_jam', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'cup', 'food_vegetable', 'photoframe', 'basket_for_clothes', 'doorjamb', 'mouse', 'door', 'table', 'light']\nTask:Chop vegetables\nresponse:\n", "output": {"thoughts": "Cut up veggies with knife", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <cutting_board>", "[GRAB] <cutting_board>", "[CLOSE] <kitchen_cabinet>", "[WALK] <table>", "[PUTBACK] <knife> <table>", "[PUTBACK] <cutting_board> <table>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <food_vegetable> <cutting_board>", "[GRAB] <knife>", "[CUT] <food_vegetable>", "[PUTOBJBACK] <knife>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'cat', 'drawing', 'hanger', 'longboard', 'floor', 'phone', 'character', 'desk', 'purse', 'stovefan', 'ceilinglamp', 'iron', 'stereo', 'window', 'oil', 'couch', 'wall', 'bench', 'toy', 'food_food', 'television', 'food_rice', 'towel_rack', 'console', 'hairbrush', 'napkin', 'table', 'vase', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I would walk into the living room. Then sit on the couch and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'toilet', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'toaster', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'love_seat', 'clothes_gloves', 'television', 'tray', 'curtain', 'food_onion', 'dresser', 'cupboard', 'instrument_guitar', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I'll watch TV", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'detergent', 'milk', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'stamp', 'food_cake', 'computer', 'piano_bench', 'mat', 'tvstand', 'faucet', 'shower', 'couch', 'clothes_hat', 'pillow', 'wall', 'powersocket', 'toy', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'mouse', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room and I lie down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'table_cloth', 'stovefan', 'laptop', 'tvstand', 'picture', 'window', 'homework', 'closetdrawer', 'couch', 'food_jam', 'wall', 'television', 'curtain', 'mouse', 'dresser', 'sink', 'instrument_guitar', 'dishwasher', 'laser_pointer']\nTask:Open window\nresponse:\n", "output": {"thoughts": "Unlock the latch at the top. Push the bottom half of the window upwards.", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'tablelamp', 'microwave', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_bacon', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'food_food', 'television', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": go to the kitchen and cook some thing to eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <oven>', '[OPEN] <oven>', '[FIND] <food_pizza>', '[GRAB] <food_pizza>', '[PUTIN] <food_pizza> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'drawing', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'mail', 'clothes_jacket', 'computer', 'ceilinglamp', 'plate', 'mat', 'picture', 'form', 'wall', 'bed', 'food_food', 'check', 'television', 'oven_mitts', 'diary', 'pencil', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'hairdryer', 'door', 'table', 'laser_pointer']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "Put food on the table, sit down, and eat the food.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <plate> <table>", "[FIND] <chair>", "[SIT] <chair>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I sit at my desk and work on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'freezer', 'walllamp', 'bathroom', 'cpuscreen', 'floor', 'character', 'desk', 'folder', 'keyboard', 'computer', 'mail', 'bookmark', 'laptop', 'ceilinglamp', 'faucet', 'shower', 'window', 'mousepad', 'closetdrawer', 'food_orange', 'pillow', 'wall', 'love_seat', 'bed', 'photoframe', 'mouse', 'electrical_outlet']\nTask:Write book\nresponse:\n", "output": {"thoughts": "Writing a book", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <electrical_outlet>", "[PLUGIN] <laptop>", "[WALK] <love_seat>", "[SIT] <love_seat>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'sponge', 'scissors', 'microwave', 'ground_coffee', 'dough', 'orchid', 'needle', 'cpuscreen', 'floor', 'ceiling', 'tea_bag', 'food_cake', 'toaster', 'stovefan', 'tvstand', 'food_orange', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'coffe_maker', 'coffee_filter', 'hairbrush', 'napkin', 'bathroom_cabinet', 'home_office', 'dishwasher', 'vase']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open the coffee maker and put a filer, grab ground coffee and put it in the coffee maker, then close the coffee maker and switch it on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[FIND] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'filing_cabinet', 'tablelamp', 'mop', 'microwave', 'chair', 'dog', 'toothbrush', 'knifeblock', 'hanger', 'ceiling', 'bedroom', 'man', 'keyboard', 'ceilinglamp', 'stereo', 'shower', 'couch', 'food_jam', 'pillow', 'pasta', 'powersocket', 'coffe_maker', 'kettle', 'pencil', 'napkin', 'table', 'scrabble', 'creditcard']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "When someone knocks I n the door walk through the hallway and greet them.", "actions": ["[RUN] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'toilet', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'envelope', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'stereo', 'mousepad', 'towel', 'wall', 'powersocket', 'photoframe', 'bed', 'toilet_paper', 'tray', 'curtain', 'doorjamb', 'dresser', 'controller', 'door', 'food_steak', 'light', 'tooth_paste']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": walk to bathroom , find shower , take a shower,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'juice', 'toilet', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'food_peanut_butter', 'food_jam', 'bowl', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'food_steak', 'light', 'dishwasher', 'table']\nTask:Grab some juice\nresponse:\n", "output": {"thoughts": "I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTOBJBACK] <juice>", "[CLOSE] <freezer>", "[FIND] <table>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_ice_cream', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'longboard', 'floor', 'bookshelf', 'computer', 'faucet', 'balanceball', 'window', 'iron', 'food_peanut_butter', 'food_jam', 'wall', 'nightstand', 'bench', 'powersocket', 'basket_for_clothes', 'novel', 'check', 'coffee_filter', 'shaving_cream', 'pencil', 'maindoor', 'sink', 'hairdryer']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter the Home office, then walk to the phone, pickup the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sponge', 'scissors', 'juice', 'food_salt', 'walllamp', 'toothbrush', 'bathroom_counter', 'shampoo', 'character', 'stamp', 'bookshelf', 'computer', 'toaster', 'stovefan', 'window', 'clothes_pants', 'pillow', 'bowl', 'food_kiwi', 'dining_room', 'oven', 'check', 'food_food', 'razor', 'ice', 'curtain', 'pencil', 'cupboard', 'sink', 'dishwasher']\nTask:Shut off alarm\nresponse:\n", "output": {"thoughts": "Shutting of the alarm clock", "actions": ["[WALK] <bedroom>", "[WALK] <alarm_clock>", "[FIND] <alarm_clock>", "[GRAB] <alarm_clock>", "[SWITCHOFF] <alarm_clock>", "[PUTOBJBACK] <alarm_clock>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'folder', 'desk', 'alcohol', 'stovefan', 'plate', 'mousepad', 'wall', 'oven', 'food_food', 'check', 'tray', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'sink', 'clothes_scarf', 'light', 'clothes_shirt', 'teeth']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk over and pick up the phone,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['spectacles', 'microwave', 'chair', 'freezer', 'walllamp', 'detergent', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'stereo', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door', 'bag', 'light', 'table']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Press the power button on the computer to turn it on", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'check', 'television', 'doorjamb', 'mouse', 'address_book', 'dresser', 'door']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Walk to Home office. Sit on the chair. Switch on the computer. Find address in addressbook. Type in the keyboard. Write mail.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <address_book>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_salt', 'chair', 'wallshelf', 'character', 'keyboard', 'food_cake', 'mop_bucket', 'stovefan', 'toaster', 'iron', 'shower', 'window', 'clothes_socks', 'closetdrawer', 'dining_room', 'broom', 'box', 'coffe_maker', 'television', 'tray', 'centerpiece', 'doorjamb', 'kettle', 'mouse', 'headset', 'cupboard', 'bag', 'table', 'light']\nTask:Shut front door\nresponse:\n", "output": {"thoughts": "Close the storm door. Lock the storm door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'scissors', 'tea', 'freezer', 'blow_dryer', 'drawing', 'knifeblock', 'cpuscreen', 'clothes_underwear', 'character', 'desk', 'toaster', 'shower', 'picture', 'mousepad', 'candle', 'dining_room', 'bench', 'oven', 'basket_for_clothes', 'bed', 'check', 'television', 'toothbrush_holder', 'curtain', 'mouse', 'sink', 'light', 'tooth_paste', 'laser_pointer']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the living room and then I walk to the light switch in there and turn the lights on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'crayon', 'mop', 'toilet', 'chair', 'bathroom', 'detergent', 'dough', 'ceiling', 'keyboard', 'stereo', 'couch', 'pillow', 'cup', 'newspaper', 'keys', 'check', 'toothbrush_holder', 'food_donut', 'after_shave', 'rag', 'mouse', 'pajamas', 'cupboard', 'bag', 'light', 'table', 'tooth_paste']\nTask:Put away keys\nresponse:\n", "output": {"thoughts": "Return keys into a bag", "actions": ["[WALK] <home_office>", "[WALK] <keys>", "[FIND] <keys>", "[GRAB] <keys>", "[FIND] <bag>", "[OPEN] <bag>", "[PUTBACK] <keys> <bag>", "[CLOSE] <bag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'dough', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'stovefan', 'mousepad', 'wall', 'nightstand', 'television', 'curtain', 'after_shave', 'mouse', 'kitchen_counter', 'clothes_dress', 'door', 'sink', 'home_office', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": find a book to read. sit down. read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sponge', 'sheets', 'pot', 'blender', 'microwave', 'chair', 'nail_polish', 'detergent', 'milk', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'clothes_socks', 'broom', 'wall', 'food_kiwi', 'powersocket', 'cup', 'photoframe', 'keys', 'newspaper', 'curtain', 'lighter', 'pencil', 'pajamas', 'clothes_dress', 'sink', 'table', 'light']\nTask:Sit in chair\nresponse:\n", "output": {"thoughts": "I sit in a chair to rest and eat.", "actions": ["[WALK] <dining_room>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'walllamp', 'drawing', 'needle', 'orchid', 'floor', 'stamp', 'keyboard', 'plate', 'faucet', 'shower', 'dish_soap', 'oil', 'wall', 'keys', 'coffe_maker', 'coffee_filter', 'pencil', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'controller', 'pajamas', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'mop', 'walllamp', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'dvd_player', 'desk', 'tvstand', 'window', 'clothes_socks', 'closetdrawer', 'oil', 'dining_room', 'wall', 'toy', 'photoframe', 'television', 'curtain', 'rag', 'doorjamb', 'console', 'kitchen_counter', 'shoes', 'dresser', 'pajamas', 'bathroom_cabinet', 'table', 'tooth_paste']\nTask:Put away shoes\nresponse:\n", "output": {"thoughts": "Pick up shoes from entry hall. Open entry way closet. Put shoes in closet. Close closet door.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[FIND] <dresser>", "[OPEN] <dresser>", "[PUTIN] <shoes> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'band-aids', 'freezer', 'bathroom', 'orchid', 'needle', 'wallshelf', 'knife', 'hanger', 'floor', 'ceiling', 'bedroom', 'keyboard', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'wall', 'oven', 'food_food', 'oven_mitts', 'doorjamb', 'pencil', 'cd', 'dresser', 'door', 'maindoor', 'bathroom_cabinet', 'light', 'clothes_skirt']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk to the bedroom and use the button on the wall, next to the door to turn on the lights.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'walllamp', 'cat', 'milk', 'orchid', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'bookshelf', 'toaster', 'ceilinglamp', 'mat', 'stereo', 'picture', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'basket_for_clothes', 'photoframe', 'doorjamb', 'door', 'light', 'table']\nTask:Look at painting\nresponse:\n", "output": {"thoughts": "Walk to the painting.  Look up at the painting.  Admire it on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'tea_bag', 'bookshelf', 'food_bacon', 'toaster', 'plate', 'tvstand', 'dish_soap', 'mousepad', 'couch', 'wall', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'food_food', 'toothbrush_holder', 'ice', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'clothes_dress', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'juice', 'tea', 'chair', 'freezer', 'bathroom', 'detergent', 'wallshelf', 'cpuscreen', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'window', 'dustpan', 'bed', 'light_bulb', 'check', 'food_food', 'curtain', 'dresser', 'door', 'napkin', 'light', 'colander', 'clothes_shirt', 'scrabble', 'dirt']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Walk up to the door.  Open the door and greet whomever is out side", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'blender', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'floor', 'bedroom', 'bookshelf', 'stovefan', 'food_peanut_butter', 'clothes_hat', 'clothes_pants', 'dining_room', 'nightstand', 'wall_clock', 'video_game_controller', 'food_donut', 'curtain', 'kettle', 'shoes', 'dresser', 'cupboard', 'hairbrush', 'controller', 'home_office', 'light', 'clothes_skirt', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the radio. Turn on the radio. Listen to the radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'glue', 'measuring_cup', 'microwave', 'chair', 'blow_dryer', 'knifeblock', 'drawing', 'orchid', 'floor', 'phone', 'bookshelf', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'wall', 'nightstand', 'photoframe', 'food_food', 'coffee_pot', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'instrument_violin', 'table', 'soap', 'tooth_paste']\nTask:Bring dirty plate to sink\nresponse:\n", "output": {"thoughts": "Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[TURNTO] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[PUTBACK] <plate> <sink>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'juice', 'food_ice_cream', 'food_oatmeal', 'microwave', 'chair', 'freezer', 'walllamp', 'ground_coffee', 'milk', 'knifeblock', 'drawing', 'knife', 'floor', 'ceiling', 'stovefan', 'toaster', 'mat', 'faucet', 'ceilinglamp', 'bowl', 'wall', 'oven', 'bench', 'cup', 'coffe_maker', 'food_donut', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Put oatmeal into a bowl.  Add milk.  Heat in a microwave.  Put pod of coffee into a coffee maker.  Activate.  Pour juice into a glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <bowl>', '[GRAB] <bowl>', '[FIND] <food_oatmeal>', '[GRAB] <food_oatmeal>', '[PUTBACK] <food_oatmeal> <bowl>', '[GRAB] <food_oatmeal>', '[PUTOBJBACK] <food_oatmeal>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[POUR] <milk> <bowl>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>', '[FIND] <microwave>', '[OPEN] <microwave>', '[PUTIN] <bowl> <microwave>', '[CLOSE] <microwave>', '[SWITCHON] <microwave>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'toilet', 'jelly', 'freezer', 'ground_coffee', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'button', 'mat', 'shower', 'picture', 'dustpan', 'food_orange', 'form', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'light_bulb', 'food_food', 'doorjamb', 'cd', 'door', 'light']\nTask:Added meat to freezer\nresponse:\n", "output": {"thoughts": "After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'mat', 'tvstand', 'faucet', 'stereo', 'picture', 'mousepad', 'couch', 'microphone', 'pillow', 'wall', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'lighter', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to living room and lie down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'dog', 'freezer', 'drawing', 'cpuscreen', 'phone', 'tea_bag', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'window', 'oil', 'food_orange', 'food_cheese', 'clothes_gloves', 'toilet_paper', 'food_food', 'television', 'coffee_filter', 'check', 'curtain', 'kettle', 'mouse', 'food_steak', 'spoon', 'instrument_guitar', 'light', 'dishwasher']\nTask:Walk to room\nresponse:\n", "output": {"thoughts": "Leave the living room and enter the hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'computer', 'mail', 'table_cloth', 'mat', 'tvstand', 'plate', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'curtain', 'after_shave', 'doorjamb', 'cards', 'mouse', 'shoes', 'sink']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I sit in the sofa and I relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'beer', 'toilet', 'chair', 'freezer', 'walllamp', 'food_butter', 'drawing', 'orchid', 'floor', 'ceiling', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'food_food', 'television', 'food_onion', 'doorjamb', 'door', 'spoon', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a can of beer, open it, sit on couch, relax", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <beer>", "[GRAB] <beer>", "[CLOSE] <freezer>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[DRINK] <beer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'juice', 'toilet', 'walllamp', 'toothbrush', 'drawing', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stamp', 'faucet', 'balanceball', 'clothes_pants', 'wall', 'powersocket', 'photoframe', 'novel', 'toothbrush_holder', 'television', 'food_donut', 'tray', 'towel_rack', 'cupboard', 'controller', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I go into the bathroom, go to the sink, get toothbrush and toothpaste. Put toothpaste on the toothbrush, brush teeth, turn on faucet, rinse brush, and return toothbrush.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['blender', 'toilet', 'chair', 'walllamp', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'ceiling', 'bedroom', 'bookshelf', 'food_bacon', 'table_cloth', 'toaster', 'laptop', 'ceilinglamp', 'floor_lamp', 'mousepad', 'couch', 'clothes_pants', 'towel', 'dining_room', 'basket_for_clothes', 'check', 'pencil', 'chessboard', 'food_steak', 'napkin', 'spoon', 'table']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Walk to the lamp.  Find the switch.  Click the switch off.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[SWITCHOFF] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'bathroom', 'ceilingfan', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'stovefan', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'cutting_board', 'bed', 'novel', 'mouthwash', 'television', 'curtain', 'towel_rack', 'home_office', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I come home and go to the sofa for turn off the tv and relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'wallshelf', 'ceiling', 'phone', 'keyboard', 'thread', 'piano_bench', 'brush', 'mat', 'ceilinglamp', 'window', 'oil', 'closetdrawer', 'mousepad', 'food_peanut_butter', 'candle', 'nightstand', 'powersocket', 'newspaper', 'novel', 'television', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'home_office', 'table', 'instrument_violin']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Twist the lamp switch until the light goes on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'toilet', 'bathroom', 'trashcan', 'ceilingfan', 'toothbrush', 'orchid', 'floor', 'phone', 'bookshelf', 'computer', 'toaster', 'shoe-shine_kit', 'mousepad', 'pillow', 'dining_room', 'box', 'basket_for_clothes', 'keys', 'razor', 'toothbrush_holder', 'coin', 'wine', 'doorjamb', 'cards', 'door', 'napkin', 'home_office', 'soap', 'tooth_paste']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <toothbrush>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'mop', 'microwave', 'detergent', 'food_butter', 'drawing', 'keyboard', 'bookshelf', 'computer', 'food_bacon', 'toaster', 'faucet', 'clothes_hat', 'oven', 'cutting_board', 'cup', 'check', 'coin', 'doorjamb', 'mouse', 'vacuum_cleaner', 'headset', 'chessboard', 'dresser', 'laundry_detergent', 'light', 'dishwasher', 'instrument_violin', 'dirt']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Get the vacuum and use it to clean my floors", "actions": ["[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[WALK] <bedroom>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'toaster', 'deck_of_cards', 'window', 'tvstand', 'mousepad', 'pillow', 'wall', 'pasta', 'light_bulb', 'check', 'mouthwash', 'wooden_spoon', 'mouse', 'door', 'table', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Turn on computer, pull up browser, browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_bread', 'toilet', 'bathroom', 'detergent', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'mail', 'ceilinglamp', 'shower', 'iron', 'closetdrawer', 'couch', 'towel', 'oven', 'coin', 'lighter', 'console', 'mouse', 'shoes', 'cd', 'dishrack', 'spoon', 'light', 'table']\nTask:Switch on lamp\nresponse:\n", "output": {"thoughts": "Press the button so the light comes on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'band-aids', 'walllamp', 'blow_dryer', 'detergent', 'bathroom', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'stereo', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'clothes_gloves', 'water', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'napkin']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <sink>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['chair', 'band-aids', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'cleaning_solution', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'photoframe', 'novel', 'rag', 'doorjamb', 'centerpiece', 'mouse', 'kitchen_counter', 'dresser', 'door', 'food_steak', 'home_office', 'dishwasher']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Gather wood cleaner and cleaning rag. Wipe baseboards and window sills with dry cloth, then spray with cleaner and buff", "actions": ["[WALK] <bedroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <wall>", "[WALK] <window>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'piano_bench', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'stereo', 'deck_of_cards', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'television', 'food_rice', 'mouse', 'door', 'table', 'laser_pointer']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to the office and sit on the chair, then work on the computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'shampoo', 'ceiling', 'button', 'deck_of_cards', 'shower', 'clothes_pants', 'food_kiwi', 'cup', 'bed', 'food_food', 'tray', 'wooden_spoon', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'spoon', 'home_office', 'light', 'clothes_skirt', 'dishwasher', 'coffee']\nTask:Grab things\nresponse:\n", "output": {"thoughts": "The robot will need some kind of hands or grabbing device to transport objects.", "actions": ["[WALK] <novel>", "[TURNTO] <novel>", "[GRAB] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'brush', 'ceilinglamp', 'mat', 'laptop', 'mousepad', 'dining_room', 'wall', 'powersocket', 'food_food', 'oven_mitts', 'kettle', 'mouse', 'cd', 'door', 'hairdryer', 'napkin', 'sink', 'table', 'light']\nTask:Wash monitor\nresponse:\n{\n    \"thoughts\": Clean the grime off of my monitor screen when not in use,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHOFF] <computer>', '[FIND] <brush>', '[TURNTO] <brush>', '[LOOKAT] <brush>', '[GRAB] <brush>', '[WIPE] <computer>', '[DROP] <brush>', '[FIND] <napkin>', '[GRAB] <napkin>', '[WASH] <napkin>', '[WIPE] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'measuring_cup', 'chair', 'walllamp', 'bathroom', 'bills', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'alcohol', 'bookmark', 'toaster', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'check', 'television', 'tray', 'cupboard', 'instrument_violin', 'teeth']\nTask:Do homework\nresponse:\n", "output": {"thoughts": "Sit on the couch in the Living room and pick u some papers", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <check>", "[GRAB] <check>", "[DROP] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'brush', 'stovefan', 'faucet', 'iron', 'couch', 'towel', 'wall', 'powersocket', 'cutting_board', 'photoframe', 'bed', 'television', 'lighter', 'console', 'dresser', 'pajamas', 'door', 'sink', 'bathroom_cabinet', 'table', 'clothes_skirt', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'scissors', 'crayon', 'toilet', 'jelly', 'walllamp', 'nail_polish', 'detergent', 'food_cake', 'alcohol', 'piano_bench', 'toaster', 'food_noodles', 'garbage_can', 'pillow', 'broom', 'oven', 'toy', 'food_vegetable', 'food_food', 'television', 'oven_mitts', 'doorjamb', 'pencil', 'headset', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'pantry']\nTask:Organize pantry\nresponse:\n", "output": {"thoughts": "Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.", "actions": ["[WALK] <dining_room>", "[WALK] <pantry>", "[FIND] <food_food>", "[GRAB] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[FIND] <garbage_can>", "[PUTBACK] <food_food> <garbage_can>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <dry_pasta>", "[GRAB] <dry_pasta>", "[TURNTO] <dry_pasta>", "[LOOKAT] <dry_pasta>", "[PUTOBJBACK] <dry_pasta>", "[FIND] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[PUSH] <food_food>", "[FIND] <food_noodles>", "[GRAB] <food_noodles>", "[TURNTO] <food_noodles>", "[LOOKAT] <food_noodles>", "[PUTOBJBACK] <food_noodles>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'bathroom', 'detergent', 'nail_polish', 'trashcan', 'food_butter', 'orchid', 'hanger', 'longboard', 'floor', 'bathroom_counter', 'remote_control', 'envelope', 'computer', 'window', 'electric_shaver', 'wall', 'cutting_board', 'bed', 'television', 'tray', 'curtain', 'after_shave', 'lighter', 'mouse', 'cupboard', 'clothes_dress', 'light', 'clothes_skirt', 'board_game']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watching tv with a remote is always fun; lay back, relax and enjoy;,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'mop', 'bathroom', 'dough', 'milk', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'cleaning_solution', 'keyboard', 'mat', 'faucet', 'window', 'food_noodles', 'closetdrawer', 'couch', 'clothes_pants', 'nightstand', 'tape', 'food_food', 'check', 'food_rice', 'wooden_spoon', 'rag', 'mouse', 'controller', 'door', 'home_office']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'freezer', 'band-aids', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'stamp', 'desk', 'stovefan', 'shoe-shine_kit', 'couch', 'pillow', 'wall', 'bench', 'coffe_maker', 'toothbrush_holder', 'television', 'curtain', 'doorjamb', 'shaving_cream', 'cupboard', 'door', 'sink', 'bathtub', 'instrument_guitar', 'light', 'home_office', 'clothes_skirt']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to room. I enter room. I find where the lightswitch is. I turn on the lights with the lightswitch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[TURNTO] <light>', '[LOOKAT] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'deck_of_cards', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'door', 'light', 'table']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Keep my keyboard vacuumed and free of debris.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'crayon', 'chair', 'freezer', 'detergent', 'food_butter', 'drawing', 'knife', 'hanger', 'floor', 'character', 'piano_bench', 'toaster', 'clothes_socks', 'food_peanut_butter', 'pillow', 'food_kiwi', 'coffe_maker', 'food_donut', 'tray', 'curtain', 'shaving_cream', 'mouse', 'kitchen_counter', 'dishrack', 'hair', 'door', 'hairdryer', 'napkin']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Turning on lights,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sheets', 'orchid', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'phone', 'cleaning_solution', 'button', 'ceilinglamp', 'mat', 'shoe-shine_kit', 'food_orange', 'pillow', 'dining_room', 'food_kiwi', 'cup', 'food_cheese', 'bed', 'clothes_gloves', 'coffe_maker', 'toilet_paper', 'oven_mitts', 'curtain', 'rag', 'console', 'pencil', 'dresser', 'clothes_scarf', 'light', 'colander']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Spray counter tops with all purpose cleaner. Using towel, wipe counter, removing all dirt and grime until clean.", "actions": ["[WALK] <dining_room>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'band-aids', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'phone', 'cleaning_solution', 'bookshelf', 'computer', 'stereo', 'oil', 'mousepad', 'wall', 'wall_clock', 'nightstand', 'shelf', 'newspaper', 'food_food', 'video_game_controller', 'coin', 'rag', 'mouse', 'pencil', 'bathtub', 'bathroom_cabinet', 'light', 'coffee', 'creditcard']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'blender', 'beer', 'microwave', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'table_cloth', 'toaster', 'plate', 'faucet', 'tvstand', 'window', 'dish_soap', 'food_orange', 'wall', 'cup', 'bed', 'food_food', 'tray', 'wooden_spoon', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'sink']\nTask:Wash dirty dishes\nresponse:\n", "output": {"thoughts": "Wet the sponge, apply soap to sponge, scrub dishes, rinse dishes, place on dish rack.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <sponge> <sink>", "[GRAB] <sponge>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'walllamp', 'bathroom', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'shoe-shine_kit', 'faucet', 'window', 'mousepad', 'couch', 'wall', 'oven', 'newspaper', 'coffe_maker', 'food_food', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'door', 'sink', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'ground_coffee', 'wallshelf', 'hanger', 'floor', 'stamp', 'desk', 'bookshelf', 'tea_bag', 'envelope', 'computer', 'mop_bucket', 'mat', 'shower', 'foundation', 'form', 'wall', 'powersocket', 'cutting_board', 'photoframe', 'novel', 'check', 'television', 'wooden_spoon', 'after_shave', 'doorjamb', 'door', 'light']\nTask:Read to child\nresponse:\n", "output": {"thoughts": "Read a book out loud to a child", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'cat', 'ceilingfan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'oil', 'couch', 'wall', 'wall_clock', 'powersocket', 'oven', 'bench', 'newspaper', 'mouthwash', 'television', 'diary', 'towel_rack', 'mouse', 'dresser', 'hairbrush', 'hairdryer', 'soap']\nTask:Check email\nresponse:\n{\n    \"thoughts\": Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[WALK] <chair>', '[SIT] <chair>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'nail_polish', 'bathroom', 'cat', 'dough', 'toothbrush', 'desk', 'button', 'stovefan', 'shower', 'faucet', 'window', 'pillow', 'nightstand', 'wall_clock', 'oven', 'cutting_board', 'novel', 'check', 'television', 'toothbrush_holder', 'lighter', 'kettle', 'chessboard', 'dresser', 'sink', 'home_office', 'clothes_skirt']\nTask:Keep cats out of room\nresponse:\n", "output": {"thoughts": "Stand in the doorway and create a loud noise so the cats will stay out.", "actions": ["[WALK] <bedroom>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'tea', 'trashcan', 'drawing', 'orchid', 'floor', 'character', 'desk', 'bookshelf', 'computer', 'toaster', 'mat', 'stovefan', 'tvstand', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'wall_clock', 'cup', 'bed', 'food_donut', 'curtain', 'towel_rack', 'sink', 'napkin', 'bathtub']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <sheets>', '[PULL] <sheets>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sheets', 'filing_cabinet', 'food_salt', 'toilet', 'band-aids', 'walllamp', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'phone', 'computer', 'mop_bucket', 'stovefan', 'tvstand', 'deck_of_cards', 'closetdrawer', 'newspaper', 'coffe_maker', 'doorjamb', 'dresser', 'cupboard', 'door', 'food_steak', 'bag', 'sink', 'table', 'dishwasher', 'bathroom_cabinet']\nTask:Put umbrella away\nresponse:\n", "output": {"thoughts": "Place the umbrella in the umbrella bucket point first.", "actions": ["[WALK] <home_office>", "[WALK] <mop_bucket>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'microwave', 'walllamp', 'bathroom', 'cat', 'trashcan', 'knifeblock', 'hanger', 'floor', 'faucet', 'shower', 'clothes_hat', 'wall', 'nightstand', 'tape', 'cup', 'food_food', 'water', 'tray', 'diary', 'pencil', 'kitchen_counter', 'headset', 'hairbrush', 'cupboard', 'hair', 'sink', 'instrument_guitar', 'soap']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <sink>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'window', 'oil', 'mousepad', 'couch', 'closetdrawer', 'food_jam', 'pillow', 'towel', 'dining_room', 'bowl', 'wall', 'nightstand', 'pasta', 'clothes_gloves', 'food_food', 'television', 'tray', 'hairbrush', 'clothes_dress']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on a couch, grab a book, open it, and scan the pages", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'detergent', 'cat', 'drawing', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'dustpan', 'pillow', 'wall', 'wall_clock', 'television', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'table', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I see my cat on the couch so I walk over, sit down and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'tea', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'folder', 'laptop', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'towel', 'wall', 'box', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'vacuum_cleaner', 'dresser', 'door', 'clothes_scarf', 'light', 'creditcard']\nTask:Sweep hallway please\nresponse:\n", "output": {"thoughts": "Vacuum all the way up and down the hallway.  Get all the dirt and crumbs out of carpet please.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'dog', 'chair', 'cat', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'phone', 'character', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'closetdrawer', 'oven', 'food_cheese', 'novel', 'toothbrush_holder', 'television', 'food_donut', 'console', 'controller', 'clothes_scarf', 'light', 'colander', 'board_game', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my cat, and I pet him,\n    \"actions\": ['[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'band-aids', 'drawing', 'floor', 'ceiling', 'desk', 'bookshelf', 'envelope', 'button', 'toaster', 'mat', 'clothes_hat', 'wall', 'powersocket', 'oven', 'toy', 'novel', 'toothbrush_holder', 'after_shave', 'doorjamb', 'pencil', 'door', 'light', 'table', 'dishwasher', 'colander']\nTask:Put toys away\nresponse:\n", "output": {"thoughts": "Pick up all the toys on the floor and put them in their correct storage bin or shelf.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blow_dryer', 'trashcan', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookmark', 'toaster', 'stovefan', 'tvstand', 'food_jam', 'pillow', 'dining_room', 'bowl', 'wall', 'powersocket', 'clothes_gloves', 'novel', 'coffe_maker', 'coffee_filter', 'tray', 'after_shave', 'curtain', 'cd', 'chessboard', 'cupboard', 'sink', 'dishwasher', 'scrabble', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[WALK] <coffee>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'measuring_cup', 'microwave', 'milk', 'knifeblock', 'bills', 'wallshelf', 'bathroom_counter', 'ceiling', 'phone', 'bedroom', 'keyboard', 'stovefan', 'laptop', 'shoe-shine_kit', 'shower', 'candle', 'clothes_hat', 'towel', 'dining_room', 'wall_clock', 'powersocket', 'newspaper', 'doorjamb', 'shaving_cream', 'vacuum_cleaner', 'dishrack', 'dishwasher', 'instrument_violin', 'creditcard']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Robot has a vacuum underneath its feat/body that will allow it to clean the surface of the floor.", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[WALK] <dining_room>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[WALK] <home_office>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <filing_cabinet>", "[PUTBACK] <vacuum_cleaner> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'form', 'pillow', 'wall', 'oven', 'box', 'newspaper', 'coffe_maker', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'pencil', 'cupboard', 'maindoor', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to living room. Walk to and sit at the sofa. Grab remote control and switch on TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'beer', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'milk', 'drawing', 'floor', 'ceiling', 'toaster', 'mat', 'shower', 'mousepad', 'garbage_can', 'food_dessert', 'wall', 'nightstand', 'powersocket', 'oven', 'food_vegetable', 'food_cheese', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'laundry_detergent', 'door', 'light']\nTask:Rotate stock in refrigerator\nresponse:\n", "output": {"thoughts": "Throw out expired items, put older items in front.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[TURNTO] <food_bread>", "[LOOKAT] <food_bread>", "[PUTOBJBACK] <food_bread>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[FIND] <garbage_can>", "[PUTBACK] <food_cheese> <garbage_can>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[PUTOBJBACK] <food_cheese>", "[FIND] <food_dessert>", "[TURNTO] <food_dessert>", "[LOOKAT] <food_dessert>", "[PUSH] <food_dessert>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <milk>", "[GRAB] <milk>", "[TURNTO] <milk>", "[LOOKAT] <milk>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'orchid', 'kitchen_cabinet', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'toaster', 'stovefan', 'faucet', 'mousepad', 'couch', 'wall', 'pasta', 'oven', 'box', 'food_food', 'water', 'tray', 'pencil', 'kitchen_counter', 'sauce_pan', 'cupboard', 'instrument_guitar', 'bathroom_cabinet']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Cook some pasta on the stove,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pasta>', '[GRAB] <pasta>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <sauce_pan>', '[POUR] <water> <sauce_pan>', '[FIND] <oven>', '[SWITCHON] <oven>', '[PUTBACK] <pasta> <sauce_pan>', '[FIND] <sauce_pan>', '[GRAB] <sauce_pan>', '[PUTBACK] <sauce_pan> <oven>']\n}\n", "input": "\nScene_Objects:['chair', 'ground_coffee', 'nail_polish', 'orchid', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'food_cake', 'bookshelf', 'mail', 'toaster', 'plate', 'homework', 'pillow', 'dining_room', 'wall', 'bed', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[GRAB] <ground_coffee>", "[PUTOBJBACK] <ground_coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'alcohol', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'iron', 'faucet', 'mousepad', 'bowl', 'wall', 'nightstand', 'oven', 'mouse', 'cupboard', 'door', 'clothes_scarf', 'bathroom_cabinet', 'table', 'instrument_guitar']\nTask:Work\nresponse:\n{\n    \"thoughts\": For doing homework walk to home office. Walk to computer. Switch on Computer. Sit in a chair and do homework in computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'food_ice_cream', 'beer', 'food_oatmeal', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'ground_coffee', 'blow_dryer', 'food_carrot', 'dough', 'milk', 'drawing', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'faucet', 'bowl', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'photoframe', 'coffe_maker', 'food_food', 'check', 'oven_mitts', 'tray', 'doorjamb', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'napkin', 'light', 'dishwasher', 'table']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Put oatmeal into a bowl.  Add milk.  Heat in a microwave.  Put pod of coffee into a coffee maker.  Activate.  Pour juice into a glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <bowl>", "[GRAB] <bowl>", "[FIND] <food_oatmeal>", "[GRAB] <food_oatmeal>", "[PUTBACK] <food_oatmeal> <bowl>", "[GRAB] <food_oatmeal>", "[PUTOBJBACK] <food_oatmeal>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <bowl>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>", "[FIND] <microwave>", "[OPEN] <microwave>", "[PUTIN] <bowl> <microwave>", "[CLOSE] <microwave>", "[SWITCHON] <microwave>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTOBJBACK] <juice>", "[CLOSE] <freezer>", "[FIND] <table>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'tvstand', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'food_jam', 'wall', 'nightstand', 'oven', 'newspaper', 'food_food', 'vacuum_cleaner', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'chair', 'bathroom', 'orchid', 'cpuscreen', 'bathroom_counter', 'desk', 'keyboard', 'button', 'piano_bench', 'ceilinglamp', 'stereo', 'floor_lamp', 'oil', 'mousepad', 'closetdrawer', 'tape', 'love_seat', 'newspaper', 'bed', 'check', 'video_game_controller', 'mouthwash', 'curtain', 'lighter', 'diary', 'doorjamb', 'dresser', 'sink']\nTask:Read\nresponse:\n", "output": {"thoughts": "Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'chair', 'freezer', 'cat', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'stereo', 'faucet', 'picture', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'cup', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'home_office', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to the bedroom and sit at my desk. I turn my radio on and listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['blender', 'milk', 'drawing', 'orchid', 'needle', 'hanger', 'bathroom_counter', 'floor', 'phone', 'computer', 'bookmark', 'toaster', 'mat', 'tvstand', 'electric_shaver', 'clothes_socks', 'closetdrawer', 'couch', 'nightstand', 'wall', 'bed', 'keys', 'photoframe', 'tray', 'centerpiece', 'dresser', 'bag', 'soap', 'clothes_shirt', 'teeth']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Climb into bed, then read the novel", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'stovefan', 'plate', 'shower', 'picture', 'faucet', 'dish_soap', 'pillow', 'wall', 'powersocket', 'cutting_board', 'tape', 'cup', 'coffe_maker', 'food_food', 'television', 'lighter', 'centerpiece', 'mouse', 'kitchen_counter', 'chessboard', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'toilet', 'freezer', 'dough', 'drawing', 'hanger', 'floor', 'ceiling', 'mat', 'shoe-shine_kit', 'shower', 'clothes_hat', 'bowl', 'nightstand', 'wall', 'powersocket', 'wall_clock', 'photoframe', 'light_bulb', 'food_food', 'coffe_maker', 'razor', 'coffee_filter', 'doorjamb', 'hairbrush', 'door', 'home_office', 'light', 'soap']\nTask:Put away groceries\nresponse:\n", "output": {"thoughts": "Place groceries in refrigerator from shopping at the grocery store.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'tablelamp', 'microwave', 'chair', 'freezer', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'stovefan', 'shower', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'cup', 'coffe_maker', 'tray', 'doorjamb', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'table', 'dishwasher', 'light']\nTask:Wipe down counter\nresponse:\n{\n    \"thoughts\": Clean counter,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <rag>', '[SQUEEZE] <rag>', '[SWITCHOFF] <faucet>', '[WALK] <kitchen_counter>', '[WIPE] <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'remote_control', 'envelope', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'pasta', 'oven', 'love_seat', 'television', 'razor', 'curtain', 'dresser', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit and watch the television", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'tablelamp', 'blender', 'tea', 'chair', 'freezer', 'drawing', 'orchid', 'hanger', 'floor', 'shampoo', 'stamp', 'bedroom', 'desk', 'computer', 'mat', 'stereo', 'mousepad', 'food_jam', 'dining_room', 'wall', 'wall_clock', 'bench', 'photoframe', 'food_food', 'dresser', 'napkin', 'home_office', 'table', 'spoon']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'chair', 'bathroom', 'drawing', 'food_egg', 'floor', 'desk', 'bookshelf', 'computer', 'mat', 'stereo', 'shower', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'cup', 'doorjamb', 'mouse', 'kitchen_counter', 'pajamas', 'controller', 'door', 'clothes_scarf', 'light', 'dishwasher', 'board_game']\nTask:Pull up carpet\nresponse:\n", "output": {"thoughts": "Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.", "actions": ["[WALK] <home_office>", "[WALK] <mat>", "[FIND] <mat>", "[TOUCH] <mat>", "[PULL] <mat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'shampoo', 'stovefan', 'mat', 'shower', 'window', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'newspaper', 'food_food', 'drinking_glass', 'water', 'curtain', 'doorjamb', 'dresser', 'food_steak', 'door', 'home_office', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'scissors', 'microwave', 'chair', 'walllamp', 'knifeblock', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'toaster', 'plate', 'faucet', 'clothes_socks', 'food_kiwi', 'wall', 'powersocket', 'clothes_gloves', 'coffe_maker', 'after_shave', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Do dishes\nresponse:\n", "output": {"thoughts": "Turn on water, rinse off dishes, put in dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[SWITCHOFF] <faucet>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'ceilingfan', 'orchid', 'bathroom_counter', 'bookshelf', 'purse', 'stovefan', 'mat', 'iron', 'electric_shaver', 'food_noodles', 'mousepad', 'couch', 'clothes_pants', 'pillow', 'powersocket', 'keys', 'coffe_maker', 'oven_mitts', 'wine', 'shaving_cream', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'home_office', 'light', 'tooth_paste']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'jelly', 'chair', 'detergent', 'food_butter', 'knifeblock', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'button', 'toaster', 'mat', 'plate', 'faucet', 'stereo', 'pillow', 'wall', 'powersocket', 'cup', 'food_food', 'diary', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'light']\nTask:Get glass of water\nresponse:\n", "output": {"thoughts": "Get a glass from a cabinet. Hold it under the sink and turn the cold water on. Fill the glass. Turn off the water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <kitchen_cabinet>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'toaster', 'deck_of_cards', 'window', 'tvstand', 'mousepad', 'pillow', 'wall', 'pasta', 'light_bulb', 'check', 'mouthwash', 'wooden_spoon', 'mouse', 'door', 'table', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Turn on computer, pull up browser, browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'freezer', 'knife', 'drawing', 'needle', 'wallshelf', 'floor', 'phone', 'ceiling', 'dvd_player', 'food_bacon', 'ceilinglamp', 'laptop', 'plate', 'shower', 'tvstand', 'tooth_paste', 'wall', 'tape', 'cup', 'novel', 'food_food', 'television', 'coffee_filter', 'hair', 'cupboard', 'table', 'board_game', 'clothes_shirt']\nTask:Eat\nresponse:\n", "output": {"thoughts": "I get my prepared dinner set on the dining table, sit down on chair, pick up utensils, eat food", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <food_food> <table>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <plate>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'toilet', 'ceilingfan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'character', 'video_game_console', 'food_cake', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'couch', 'wall', 'powersocket', 'tape', 'cup', 'video_game_controller', 'television', 'mouthwash', 'curtain', 'hairdryer', 'table', 'coffee']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'beer', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'food_orange', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Make toast\nresponse:\n", "output": {"thoughts": "Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'dog', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'mat', 'faucet', 'shower', 'couch', 'wall', 'powersocket', 'newspaper', 'keys', 'wooden_spoon', 'doorjamb', 'centerpiece', 'kitchen_counter', 'door', 'home_office', 'light', 'shoe_rack']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Switch on light in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[SWITCHON] <light>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'microwave', 'freezer', 'blow_dryer', 'ground_coffee', 'cat', 'floor', 'legs_both', 'face', 'toaster', 'laptop', 'homework', 'closetdrawer', 'towel', 'feet_both', 'arms_both', 'basket_for_clothes', 'newspaper', 'photoframe', 'coffe_maker', 'mouthwash', 'centerpiece', 'doorjamb', 'headset', 'dresser', 'cupboard', 'napkin', 'soap', 'clothes_shirt']\nTask:Try yourself off\nresponse:\n", "output": {"thoughts": "After your shower, go into your bedroom and use a towel to dry water off of yourself before dressing.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <towel>", "[GRAB] <towel>", "[CLOSE] <dresser>", "[FIND] <face>", "[WIPE] <face>", "[FIND] <arms_both>", "[WIPE] <arms_both>", "[FIND] <legs_both>", "[WIPE] <legs_both>", "[FIND] <feet_both>", "[WIPE] <feet_both>", "[FIND] <basket_for_clothes>", "[PUTBACK] <towel> <basket_for_clothes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'ground_coffee', 'detergent', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'stovefan', 'toaster', 'faucet', 'shower', 'closetdrawer', 'candle', 'dining_room', 'wall', 'oven', 'photoframe', 'bed', 'coffe_maker', 'coffee_filter', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'table']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Make coffee in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[PLUGIN] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'measuring_cup', 'band-aids', 'knifeblock', 'cpuscreen', 'dvd_player', 'folder', 'desk', 'bookmark', 'stereo', 'faucet', 'dustpan', 'microphone', 'oven', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'television', 'food_rice', 'pencil', 'shredder', 'controller', 'home_office', 'light', 'table', 'scrabble', 'electrical_outlet']\nTask:Shred receipts\nresponse:\n", "output": {"thoughts": "Take old receipts and put through the shredder", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[FIND] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <check>", "[GRAB] <check>", "[CLOSE] <filing_cabinet>", "[FIND] <electrical_outlet>", "[FIND] <shredder>", "[PLUGIN] <shredder>", "[SWITCHON] <shredder>", "[PUTBACK] <check> <shredder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'chair', 'drawing', 'bills', 'knife', 'floor', 'ceilinglamp', 'stereo', 'window', 'couch', 'food_jam', 'form', 'towel', 'dining_room', 'wall', 'basket_for_clothes', 'photoframe', 'novel', 'food_food', 'check', 'television', 'oven_mitts', 'razor', 'tray', 'pencil', 'dresser', 'pajamas', 'table', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit and read,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[WALK] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'jelly', 'walllamp', 'bathroom', 'knife', 'wallshelf', 'floor', 'ceiling', 'bookmark', 'ceilinglamp', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'television', 'toothbrush_holder', 'wooden_spoon', 'kettle', 'shoes', 'cupboard', 'conditioner', 'clothes_dress', 'door', 'napkin', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up a book and open to the last page you were on. Read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'food_cake', 'thread', 'computer', 'laptop', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'pasta', 'television', 'tray', 'doorjamb', 'mouse', 'headset', 'chessboard', 'cupboard', 'door']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Sit in my deck chair with my laptop, find a good movie on You Tube to watch, plug in earphones and listen to movie quietly so I don't disturb my upstairs neighbor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <headset>', '[GRAB] <headset>', '[PUTON] <headset>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'scissors', 'juice', 'food_salt', 'microwave', 'chair', 'food_butter', 'toothbrush', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'button', 'brush', 'ceilinglamp', 'plate', 'wall', 'tape', 'food_food', 'check', 'food_donut', 'wooden_spoon', 'rag', 'door', 'clothes_scarf', 'napkin', 'table', 'light', 'dishwasher']\nTask:Eat\nresponse:\n", "output": {"thoughts": "Serve food on dinning table in appropriate utensils, sit on the chair, eat", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[TURNTO] <table>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <plate>", "[PUTBACK] <food_food> <plate>", "[GRAB] <food_food>", "[PUTOBJBACK] <food_food>", "[FIND] <fork>", "[GRAB] <fork>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'microwave', 'freezer', 'walllamp', 'detergent', 'toothbrush', 'orchid', 'drawing', 'clothes_underwear', 'floor', 'thread', 'mat', 'closetdrawer', 'wall', 'nightstand', 'toy', 'standingmirror', 'light_bulb', 'novel', 'check', 'tray', 'cards', 'mouse', 'pencil', 'vacuum_cleaner', 'headset', 'maindoor', 'napkin', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the Living Room. Look at the Cabinet. Find a Textbook. Grab the textbook. Open the Textbook. Read the textbook.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[FIND] <filing_cabinet>', '[TURNTO] <filing_cabinet>', '[LOOKAT] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'chair', 'freezer', 'walllamp', 'bathroom', 'woman', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'child', 'powersocket', 'check', 'coffe_maker', 'television', 'curtain', 'pencil', 'dresser', 'table']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sitting and watching  a movie with family", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TURNTO] <child>", "[LOOKAT] <child>", "[GREET] <child>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'juice', 'cd_player', 'toilet', 'freezer', 'trashcan', 'bedroom', 'remote_control', 'button', 'clothes_jacket', 'toaster', 'laptop', 'ceilinglamp', 'plate', 'homework', 'dining_room', 'broom', 'bench', 'tape', 'keys', 'coffe_maker', 'television', 'towel_rack', 'hair', 'cupboard', 'maindoor', 'sink', 'clothes_skirt', 'vase']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[PLUGIN] <cd_player>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['food_bread', 'toilet', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'ceiling', 'cleaning_solution', 'mop_bucket', 'table_cloth', 'faucet', 'shower', 'iron', 'clothes_socks', 'pillow', 'wall', 'food_food', 'mirror', 'wine', 'doorjamb', 'coffee_pot', 'paper_towel', 'cupboard', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'dishwasher', 'scrabble']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[CLOSE] <bathroom_cabinet>", "[WALK] <mirror>", "[FIND] <cleaning_solution>", "[RINSE] <cleaning_solution>", "[WIPE] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'chair', 'bathroom_counter', 'floor', 'laptop', 'faucet', 'shower', 'towel', 'wall', 'wall_clock', 'nightstand', 'cup', 'bed', 'photoframe', 'water', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go into the bathroom and go over to the sink. Grab soap and turn on the water. Wet your hands and rub them together so that the soap lathers. Rub the soap all over your hands thoroughly. Rinse off the soap with water and turn off the sink. Dry your hands with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <hands_both>', '[POUR] <water> <hands_both>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'band-aids', 'walllamp', 'detergent', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'video_game_console', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'plate', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'photoframe', 'video_game_controller', 'television', 'curtain', 'dresser', 'table', 'dishwasher']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on game.  Use controller", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[PLUGIN] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'water_glass', 'pot', 'tablelamp', 'microwave', 'drawing', 'knife', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'toaster', 'stovefan', 'plate', 'shower', 'dish_soap', 'mousepad', 'wall', 'oven', 'photoframe', 'coffe_maker', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'spoon', 'coffee_cup', 'dishwasher', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and put plates, cutlery and glasses into the dishwasher, put the soap, close the machine and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <dishwasher>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <dishwasher>', '[FIND] <pot>', '[GRAB] <pot>', '[PUTBACK] <pot> <dishwasher>', '[FIND] <pot>', '[GRAB] <pot>', '[PUTBACK] <pot> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[PUTBACK] <water_glass> <dishwasher>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[PUTBACK] <water_glass> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'microwave', 'chair', 'wallshelf', 'hanger', 'floor', 'phone', 'food_bacon', 'piano_bench', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'powersocket', 'television', 'coin', 'food_donut', 'ice', 'rag', 'pencil', 'shoes', 'dresser', 'clothes_dress', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Hang up jacket\nresponse:\n", "output": {"thoughts": "I put clothes on hangers and hang them in the closet", "actions": ["[WALK] <bedroom>", "[WALK] <clothes_dress>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[WALK] <dresser>", "[FIND] <hanger>", "[OPEN] <dresser>", "[PUTIN] <clothes_dress> <dresser>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[WALK] <dresser>", "[PUTIN] <clothes_dress> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'fork', 'pot', 'sheets', 'food_salt', 'detergent', 'food_butter', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'shampoo', 'desk', 'food_cake', 'mat', 'faucet', 'food_kiwi', 'wall', 'powersocket', 'toy', 'novel', 'food_food', 'curtain', 'dresser', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Going to the sink and scrubbing hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'freezer', 'ground_coffee', 'milk', 'orchid', 'ceiling', 'phone', 'character', 'keyboard', 'food_bacon', 'stovefan', 'ceilinglamp', 'tvstand', 'closetdrawer', 'couch', 'dining_room', 'bowl', 'nightstand', 'pasta', 'food_cheese', 'bed', 'novel', 'check', 'food_food', 'coin', 'tray', 'doorjamb', 'diary', 'table']\nTask:Look out window\nresponse:\n", "output": {"thoughts": "I head into my child's room and look out the window to view the neighborhood from a different area in the house.", "actions": ["[WALK] <bedroom>", "[WALK] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'drawing', 'hanger', 'floor', 'phone', 'ceilinglamp', 'plate', 'iron', 'candle', 'pillow', 'wall', 'bench', 'powersocket', 'oven', 'food_food', 'video_game_controller', 'check', 'wooden_spoon', 'after_shave', 'doorjamb', 'shoes', 'dresser', 'cupboard', 'clothes_dress', 'door', 'conditioner', 'sink', 'table', 'light']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": Put food on the table, sit down, and eat the food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[PUTBACK] <plate> <table>', '[FIND] <chair>', '[SIT] <chair>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['food_bread', 'tablelamp', 'microwave', 'chair', 'walllamp', 'floor', 'ceiling', 'character', 'bedroom', 'stovefan', 'ceilinglamp', 'faucet', 'oil', 'closetdrawer', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'nightstand', 'cutting_board', 'photoframe', 'television', 'mouse', 'pencil', 'hairbrush', 'pajamas', 'bathroom_cabinet', 'light', 'teeth']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Bring book, lie on the sofa, read book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[LIE] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'ceilingfan', 'drawing', 'food_egg', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'television', 'food_rice', 'curtain', 'kitchen_counter', 'hair', 'home_office', 'table', 'light', 'dishwasher', 'dirt']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'drawing', 'needle', 'cpuscreen', 'bills', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'mousepad', 'tooth_paste', 'couch', 'pillow', 'towel', 'wall', 'nightstand', 'powersocket', 'television', 'coin', 'razor', 'doorjamb', 'centerpiece', 'mouse', 'door', 'light', 'board_game']\nTask:Re arrange office\nresponse:\n", "output": {"thoughts": "Walk in.Turn on light.Move furniture. Change the room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[PUSH] <desk>", "[WALK] <chair>", "[PUSH] <chair>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'crayon', 'freezer', 'nail_polish', 'food_carrot', 'milk', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'cup', 'bed', 'food_food', 'doorjamb', 'mouse', 'pajamas', 'controller', 'door', 'spoon', 'napkin', 'light', 'dirt']\nTask:Get glass of milk\nresponse:\n{\n    \"thoughts\": Get an empty glass. Take milk from refrigerator and open it. Pour milk into glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[POUR] <milk> <cup>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'food_salt', 'tea', 'chair', 'food_carrot', 'drawing', 'floor', 'ceiling', 'phone', 'dvd_player', 'bookshelf', 'stovefan', 'mat', 'mousepad', 'food_noodles', 'dining_room', 'wall', 'powersocket', 'oven', 'cup', 'check', 'rag', 'doorjamb', 'cd', 'hairbrush', 'dresser', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Hit light switch", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'shoe-shine_kit', 'balanceball', 'picture', 'ceilinglamp', 'mousepad', 'wall', 'nightstand', 'newspaper', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I need to finish some work on my computer in my home office to meet a deadline for tomorrow.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[TYPE] <keyboard>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'knifeblock', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'mat', 'tvstand', 'faucet', 'couch', 'pillow', 'wall', 'powersocket', 'bed', 'clothes_gloves', 'novel', 'toilet_paper', 'doorjamb', 'coffee_pot', 'mouse', 'kitchen_counter', 'door', 'napkin', 'bathroom_cabinet', 'light']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Hit the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'tea', 'chair', 'walllamp', 'ground_coffee', 'bills', 'orchid', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'mail', 'ceilinglamp', 'mat', 'iron', 'stereo', 'faucet', 'clothes_socks', 'clothes_hat', 'basket_for_clothes', 'food_rice', 'curtain', 'lighter', 'doorjamb', 'towel_rack', 'vacuum_cleaner', 'dishrack', 'cd', 'napkin', 'home_office']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[TOUCH] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'detergent', 'knife', 'bills', 'bathroom_counter', 'phone', 'folder', 'bedroom', 'character', 'computer', 'toaster', 'stovefan', 'mat', 'window', 'mousepad', 'pillow', 'oven', 'check', 'coffe_maker', 'razor', 'after_shave', 'mouse', 'kitchen_counter', 'shoes', 'dresser', 'pajamas', 'sink', 'bag']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "I get dressed in my bedroom in the mornings.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'food_ice_cream', 'walllamp', 'bathroom', 'ceilingfan', 'trashcan', 'woman', 'drawing', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'bedroom', 'man', 'ceilinglamp', 'window', 'couch', 'wall', 'bench', 'cup', 'newspaper', 'novel', 'check', 'television', 'towel_rack', 'clothes_dress', 'table', 'clothes_skirt', 'creditcard']\nTask:Hang with friends\nresponse:\n{\n    \"thoughts\": Sit with friends. Talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[TURNTO] <man>', '[LOOKAT] <man>', '[GREET] <man>', '[FIND] <woman>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <man>', '[LOOKAT] <man>', '[GREET] <man>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'band-aids', 'bathroom', 'cat', 'orchid', 'wallshelf', 'hanger', 'floor', 'phone', 'character', 'bedroom', 'purse', 'computer', 'table_cloth', 'tvstand', 'powersocket', 'toy', 'clothes_gloves', 'coffe_maker', 'razor', 'food_onion', 'centerpiece', 'headset', 'chessboard', 'cupboard', 'door', 'home_office', 'table', 'laser_pointer']\nTask:Open door\nresponse:\n", "output": {"thoughts": "I walk up to the door.  I use my hand to turn the door knob to the left.  I pull the door open with the same hand I used to turn the door knob", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'faucet', 'picture', 'window', 'dustpan', 'towel', 'wall', 'nightstand', 'photoframe', 'keys', 'toothbrush_holder', 'food_donut', 'mouse', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Cleaning hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'ground_coffee', 'woman', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'shampoo', 'floor', 'ceiling', 'bookshelf', 'food_noodles', 'closetdrawer', 'mousepad', 'couch', 'towel', 'wall', 'food_food', 'television', 'kettle', 'mouse', 'cards', 'vacuum_cleaner', 'dresser', 'door', 'hairdryer', 'dishwasher', 'clothes_shirt']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet people into my home", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'fork', 'filing_cabinet', 'chair', 'cat', 'drawing', 'hanger', 'floor', 'character', 'tea_bag', 'mat', 'picture', 'electric_shaver', 'couch', 'food_jam', 'towel', 'dining_room', 'wall', 'oven', 'coffe_maker', 'tray', 'curtain', 'doorjamb', 'diary', 'cards', 'cupboard', 'door', 'bathroom_cabinet', 'light', 'tooth_paste']\nTask:Read book\nresponse:\n{\n    \"thoughts\": pick up  the book open the book and read,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'character', 'computer', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'window', 'deck_of_cards', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'food_kiwi', 'nightstand', 'wall', 'cup', 'coffe_maker', 'television', 'food_rice', 'curtain', 'dresser', 'napkin', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on my couch and watch TV in my living room", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'character', 'tea_bag', 'computer', 'table_cloth', 'toaster', 'mat', 'tvstand', 'stovefan', 'electric_shaver', 'couch', 'pillow', 'wall', 'basket_for_clothes', 'newspaper', 'coffee_table', 'photoframe', 'video_game_controller', 'television', 'doorjamb', 'dresser', 'cupboard', 'dishwasher']\nTask:Read magazine\nresponse:\n{\n    \"thoughts\": Get magazine off coffee table. Sit in arm chair. Open magazine to first page. Read until finished.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <newspaper>', '[PUTBACK] <newspaper> <coffee_table>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'freezer', 'ground_coffee', 'drawing', 'knifeblock', 'clothes_underwear', 'hanger', 'floor', 'desk', 'stamp', 'bookshelf', 'food_cake', 'table_cloth', 'bookmark', 'mat', 'clothes_hat', 'wall', 'pasta', 'powersocket', 'toy', 'box', 'food_cheese', 'light_bulb', 'doorjamb', 'lighter', 'kitchen_counter', 'door', 'light']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Enter inside kid's room, begin picking up toys from the floor and putting them in a toy box, leave once once.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'box', 'novel', 'toilet_paper', 'television', 'mouse', 'dresser', 'door', 'food_steak', 'clothes_skirt', 'tooth_paste']\nTask:Read book\nresponse:\n{\n    \"thoughts\": \"walk to entrance\" \"walk to living room\" \"look for book\" \"find book\" \"pick up book\" \"look for chair\" \"find chair\" \"sit in chair\" \"read book\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[FIND] <novel>', '[FIND] <chair>', '[TURNTO] <chair>', '[LOOKAT] <chair>', '[SIT] <chair>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'sponge', 'crayon', 'microwave', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'floor', 'bookshelf', 'button', 'stovefan', 'mat', 'stereo', 'mousepad', 'clothes_hat', 'bowl', 'nightstand', 'wall', 'bed', 'photoframe', 'light_bulb', 'food_food', 'television', 'kitchen_counter', 'dresser', 'door', 'sink', 'clothes_skirt', 'scrabble']\nTask:Breakfast\nresponse:\n", "output": {"thoughts": "I have breakfast in bed, watching tv.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[GRAB] <food_food>", "[WALK] <bedroom>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <bed>", "[LIE] <bed>", "[TURNTO] <television>", "[WATCH] <television>", "[EAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'plate', 'stereo', 'mousepad', 'wall', 'bench', 'television', 'wine', 'rag', 'lighter', 'shaving_cream', 'mouse', 'hairbrush', 'door', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": walk to living room, look at computer, find chair, sit in chair, grab keyboard, grab mouse, type using keyboard and mouse,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_ice_cream', 'beer', 'crayon', 'chair', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'phone', 'desk', 'bedroom', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'cup', 'bed', 'keys', 'food_food', 'television', 'food_onion', 'doorjamb', 'mouse', 'home_office']\nTask:Lock door\nresponse:\n", "output": {"thoughts": "Turn the lock.", "actions": ["[WALK] <bedroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'walllamp', 'bathroom', 'cat', 'ceilingfan', 'knifeblock', 'phone', 'character', 'brush', 'toaster', 'mat', 'faucet', 'ceilinglamp', 'shower', 'window', 'couch', 'wall_clock', 'cup', 'newspaper', 'food_onion', 'lighter', 'towel_rack', 'mouse', 'kitchen_counter', 'headset', 'cupboard', 'table', 'vase', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Every time this animal is seen, you must approach it and show compassion for it by petting it in a gentle manner.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <bedroom>', '[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[FIND] <brush>', '[GRAB] <brush>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'chair', 'jelly', 'knifeblock', 'floor', 'ceiling', 'character', 'alcohol', 'toaster', 'plate', 'faucet', 'drying_rack', 'window', 'tvstand', 'dish_soap', 'closetdrawer', 'couch', 'towel', 'pillow', 'food_kiwi', 'wall', 'toy', 'novel', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'instrument_guitar', 'shoe_rack']\nTask:Do dishes\nresponse:\n", "output": {"thoughts": "Wash dirty dishes, dry them and then put them away", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sink>", "[PUTOBJBACK] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[FIND] <drying_rack>", "[PUTBACK] <plate> <drying_rack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <drying_rack>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'detergent', 'ceilingfan', 'food_carrot', 'milk', 'drawing', 'knifeblock', 'food_egg', 'floor', 'phone', 'desk', 'bookshelf', 'envelope', 'ceilinglamp', 'window', 'mousepad', 'couch', 'wall', 'food_cheese', 'novel', 'television', 'food_rice', 'after_shave', 'towel_rack', 'kitchen_counter', 'comb', 'table', 'dishwasher', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and find the sofa. I sit on the sofa and get comfortable.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'chair', 'dog', 'ground_coffee', 'knife', 'needle', 'bills', 'bathroom_counter', 'floor', 'phone', 'bookshelf', 'bookmark', 'ceilinglamp', 'faucet', 'shower', 'picture', 'food_noodles', 'clothes_socks', 'mousepad', 'pillow', 'wall', 'photoframe', 'food_food', 'tray', 'kitchen_counter', 'sink', 'bag', 'home_office', 'table']\nTask:Sit\nresponse:\n", "output": {"thoughts": "We sit at the dining room table", "actions": ["[WALK] <dining_room>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'bathroom', 'detergent', 'orchid', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'purse', 'computer', 'table_cloth', 'bookmark', 'stovefan', 'stereo', 'oil', 'closetdrawer', 'coffe_maker', 'television', 'food_donut', 'food_rice', 'tray', 'centerpiece', 'doorjamb', 'pencil', 'maindoor', 'sink', 'home_office', 'table', 'clothes_skirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to living room. Find book. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'food_carrot', 'hanger', 'phone', 'desk', 'keyboard', 'computer', 'stovefan', 'deck_of_cards', 'faucet', 'window', 'closetdrawer', 'microphone', 'pillow', 'box', 'bed', 'photoframe', 'check', 'food_onion', 'diary', 'pencil', 'shoes', 'dishrack', 'dresser', 'laundry_detergent', 'spoon', 'soap', 'tooth_paste', 'shoe_rack']\nTask:Take off shoes\nresponse:\n", "output": {"thoughts": "Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <shoes> <dresser>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[PUTIN] <shoes> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'shoe-shine_kit', 'mousepad', 'pillow', 'wall', 'powersocket', 'cup', 'bed', 'wooden_spoon', 'coffee_pot', 'mouse', 'kitchen_counter', 'door', 'home_office', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I walk to the Home Office then find the computer and sit in the chair. I switch on the computer and start typing the email. I switch the computer off.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_pizza', 'tablelamp', 'beer', 'freezer', 'blow_dryer', 'drawing', 'needle', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'mat', 'mousepad', 'form', 'wall', 'powersocket', 'food_cheese', 'food_food', 'check', 'doorjamb', 'mouse', 'chessboard', 'dresser', 'laundry_detergent', 'door', 'hairdryer', 'light', 'board_game']\nTask:Put clothes away\nresponse:\n", "output": {"thoughts": "I take clean clothes and either place them in dresser drawers or hang them in a closet.", "actions": ["[WALK] <bedroom>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'chair', 'freezer', 'walllamp', 'food_butter', 'bills', 'cpuscreen', 'bathroom_counter', 'phone', 'ceiling', 'character', 'bedroom', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'floor_lamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'oven', 'cup', 'photoframe', 'television', 'kitchen_counter', 'dresser', 'bathroom_cabinet', 'table', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk over to light switch, flip the switch to the on position,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'ground_coffee', 'walllamp', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'character', 'computer', 'piano_bench', 'stovefan', 'toaster', 'tvstand', 'faucet', 'oil', 'pillow', 'wall', 'oven', 'powersocket', 'food_food', 'water', 'tray', 'after_shave', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Get my coffee pot going", "actions": ["[WALK] <dining_room>", "[WALK] <coffee_pot>", "[FIND] <coffee_pot>", "[GRAB] <coffee_pot>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffee_pot>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffee_pot>", "[GRAB] <ground_coffee>", "[PUTOBJBACK] <ground_coffee>", "[WALK] <oven>", "[PUTBACK] <coffee_pot> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'knifeblock', 'knife', 'orchid', 'floor', 'ceiling', 'phone', 'keyboard', 'mail', 'stovefan', 'ceilinglamp', 'deck_of_cards', 'stereo', 'pillow', 'dining_room', 'wall', 'cup', 'bed', 'novel', 'coffe_maker', 'doorjamb', 'kitchen_counter', 'cupboard', 'controller', 'laundry_detergent', 'hairdryer', 'sink', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Making coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'beer', 'measuring_cup', 'band-aids', 'woman', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'phone', 'ceiling', 'character', 'desk', 'man', 'thread', 'clothes_jacket', 'tvstand', 'stereo', 'faucet', 'window', 'shower', 'couch', 'form', 'cutting_board', 'novel', 'television', 'bathroom_cabinet', 'table', 'dishwasher', 'instrument_violin']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'tablelamp', 'crayon', 'microwave', 'freezer', 'bathroom', 'trashcan', 'ceilingfan', 'knifeblock', 'clothes_underwear', 'longboard', 'shampoo', 'dvd_player', 'folder', 'mail', 'bookmark', 'toaster', 'shower', 'homework', 'couch', 'bed', 'clothes_gloves', 'towel_rack', 'mouse', 'pencil', 'cupboard', 'maindoor', 'table', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'tablelamp', 'chair', 'walllamp', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'man', 'food_bacon', 'stovefan', 'ceilinglamp', 'brush', 'mat', 'oil', 'couch', 'food_jam', 'pillow', 'wall', 'nightstand', 'newspaper', 'novel', 'coffe_maker', 'television', 'mouse', 'chessboard', 'soap', 'instrument_violin', 'teeth']\nTask:Have conversation with boyfriend\nresponse:\n", "output": {"thoughts": "Sit on sofa, talk", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'microwave', 'toilet', 'dog', 'freezer', 'drawing', 'knifeblock', 'longboard', 'floor', 'folder', 'computer', 'ceilinglamp', 'window', 'couch', 'wall', 'nightstand', 'bench', 'cup', 'bed', 'clothes_gloves', 'novel', 'television', 'curtain', 'towel_rack', 'cupboard', 'sink', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I'm tired and want to relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'microwave', 'toilet', 'jelly', 'freezer', 'band-aids', 'drawing', 'knifeblock', 'clothes_underwear', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'powersocket', 'cup', 'basket_for_clothes', 'photoframe', 'food_food', 'water', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Get drink\nresponse:\n", "output": {"thoughts": "I go to the fridge and open the door. I look for the water pitcher, and take it out, making sure to close the fridge door behind me. Then I take a pint glass out of the upper cabinet over the sink and use the water pitcher to pour a full glass of water. Then I return the pitcher to the fridge and shut the door again, and take my glass of water out of the kitche.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <kitchen_cabinet>", "[WALK] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[OPEN] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'walllamp', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'thread', 'bookshelf', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'clothes_socks', 'wall', 'pasta', 'powersocket', 'cutting_board', 'cup', 'clothes_gloves', 'novel', 'television', 'rag', 'doorjamb', 'shoes', 'door', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'sponge', 'mop', 'chair', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'window', 'mousepad', 'clothes_socks', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'pasta', 'nightstand', 'light_bulb', 'coffe_maker', 'television', 'wooden_spoon', 'mouse', 'shoes', 'vacuum_cleaner', 'sink']\nTask:Check email\nresponse:\n", "output": {"thoughts": "Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[WALK] <chair>", "[SIT] <chair>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'laptop', 'faucet', 'mousepad', 'candle', 'microphone', 'food_orange', 'wall', 'photoframe', 'check', 'mouthwash', 'after_shave', 'doorjamb', 'cupboard', 'pajamas', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'light', 'soap', 'clothes_shirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go into the bathroom and turn on the light. I walk to the sink and turn on the water. I grab the soap and wash my hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[PUTOBJBACK] <soap>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'jelly', 'walllamp', 'detergent', 'knifeblock', 'orchid', 'floor', 'ceiling', 'toaster', 'plate', 'faucet', 'dish_soap', 'wall', 'nightstand', 'powersocket', 'oven', 'food_food', 'television', 'coffee_filter', 'wine', 'curtain', 'rag', 'doorjamb', 'shaving_cream', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'bag']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Run hot water into tub of dirty dishes in the sink.  Add liquid soap.  Scrub dishes with wash rag. Rinse and place on drying rack.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sink>", "[PUTOBJBACK] <dish_soap>", "[FIND] <rag>", "[GRAB] <rag>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_pizza', 'tablelamp', 'blender', 'food_salt', 'blow_dryer', 'trashcan', 'toothbrush', 'cpuscreen', 'hanger', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'purse', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'microphone', 'oven', 'food_food', 'check', 'tray', 'dresser', 'napkin', 'bathroom_cabinet', 'clothes_shirt']\nTask:Leave home\nresponse:\n{\n    \"thoughts\": Put on shoes. Put on coat. Open door and exit through it. Lock door behind you.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'jelly', 'freezer', 'detergent', 'woman', 'orchid', 'drawing', 'phone', 'character', 'man', 'bookshelf', 'toaster', 'ceilinglamp', 'tvstand', 'stereo', 'food_orange', 'nightstand', 'box', 'food_cheese', 'coffe_maker', 'food_food', 'check', 'video_game_controller', 'centerpiece', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'sink', 'table', 'tooth_paste']\nTask:Greet people\nresponse:\n", "output": {"thoughts": "Say HI to those that come in.", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'walllamp', 'knife', 'food_egg', 'floor', 'food_cake', 'toaster', 'plate', 'tvstand', 'window', 'dish_soap', 'wall', 'nightstand', 'oven', 'cutting_board', 'coffe_maker', 'check', 'razor', 'television', 'food_donut', 'tray', 'centerpiece', 'doorjamb', 'dishrack', 'food_steak', 'spoon', 'light', 'dishwasher', 'colander']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <dishrack>', '[PUTBACK] <fork> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishrack>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishrack>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['blender', 'freezer', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'brush', 'tvstand', 'stereo', 'window', 'closetdrawer', 'towel', 'wall', 'powersocket', 'tape', 'food_food', 'television', 'curtain', 'doorjamb', 'vacuum_cleaner', 'dresser', 'door', 'clothes_scarf', 'dishwasher', 'board_game']\nTask:vacuum carpet\nresponse:\n", "output": {"thoughts": "I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'longboard', 'character', 'bedroom', 'computer', 'faucet', 'homework', 'clothes_socks', 'food_noodles', 'dining_room', 'photoframe', 'check', 'television', 'after_shave', 'rag', 'lighter', 'towel_rack', 'kitchen_counter', 'dresser', 'clothes_dress', 'home_office', 'light', 'laser_pointer', 'scrabble']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office. I find the telephone. I pick up the telephone handset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'chair', 'bathroom', 'toothbrush', 'drawing', 'cpuscreen', 'floor', 'keyboard', 'piano_bench', 'ceilinglamp', 'oil', 'form', 'food_orange', 'wall', 'nightstand', 'oven', 'photoframe', 'check', 'food_food', 'television', 'coffee_filter', 'tray', 'mouse', 'cd', 'cupboard', 'door', 'table', 'light', 'creditcard']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'microwave', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'toaster', 'stovefan', 'deck_of_cards', 'shower', 'pillow', 'wall', 'powersocket', 'cup', 'coffe_maker', 'check', 'wine', 'food_onion', 'curtain', 'lighter', 'kitchen_counter', 'cupboard', 'hair', 'controller', 'table', 'dishwasher', 'colander']\nTask:Take dishes out of dishwasher\nresponse:\n{\n    \"thoughts\": Take clean dishes out of the dishwasher and put away,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[CLOSE] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'ground_coffee', 'walllamp', 'milk', 'cpuscreen', 'ceiling', 'phone', 'keyboard', 'stovefan', 'plate', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'wine', 'tray', 'food_onion', 'mouse', 'kitchen_counter', 'headset', 'comb', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Get out dish\nresponse:\n", "output": {"thoughts": "Open cupboard.  Get out the dish.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'faucet', 'window', 'oil', 'couch', 'wall', 'bench', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'after_shave', 'towel_rack', 'kitchen_counter', 'bathtub', 'sink', 'light']\nTask:Open bathroom window\nresponse:\n{\n    \"thoughts\": Make sure bathroom window is unlocked. Lift up bathroom window so it is open.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <window>', '[FIND] <window>', '[OPEN] <window>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_salt', 'cpuscreen', 'clothes_underwear', 'floor', 'bedroom', 'food_cake', 'computer', 'stovefan', 'ceilinglamp', 'oil', 'couch', 'powersocket', 'cutting_board', 'food_cheese', 'photoframe', 'food_food', 'razor', 'oven_mitts', 'coffee_pot', 'mouse', 'shoes', 'dishrack', 'chessboard', 'cupboard', 'controller', 'food_steak', 'hairdryer', 'spoon']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Clean the welcome mat", "actions": ["[WALK] <home_office>", "[WALK] <vacuum_cleaner>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'food_ice_cream', 'mop', 'microwave', 'detergent', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'candle', 'wall', 'bed', 'television', 'curtain', 'dresser', 'maindoor', 'table', 'dishwasher', 'vase', 'coffee']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to livingroom, sit on couch, grab remote and turn on tv.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'measuring_cup', 'microwave', 'jelly', 'ground_coffee', 'drawing', 'knifeblock', 'bathroom_counter', 'character', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'brush', 'faucet', 'picture', 'mat', 'closetdrawer', 'powersocket', 'oven', 'cup', 'basket_for_clothes', 'clothes_gloves', 'novel', 'television', 'kettle', 'cupboard', 'sink', 'spoon', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the book shelf, get a book, open the book and read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'tablelamp', 'crayon', 'tea', 'mop', 'ground_coffee', 'detergent', 'drawing', 'floor', 'phone', 'ceilinglamp', 'laptop', 'shower', 'iron', 'window', 'oil', 'couch', 'pillow', 'wall', 'oven', 'cup', 'coffe_maker', 'toilet_paper', 'television', 'door', 'sink', 'table', 'scrabble', 'dirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Bring book, lie on the sofa, read book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[LIE] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_pizza', 'blender', 'tea', 'toilet', 'chair', 'freezer', 'bathroom', 'woman', 'knifeblock', 'drawing', 'wallshelf', 'floor', 'phone', 'character', 'desk', 'man', 'button', 'deck_of_cards', 'mousepad', 'clothes_pants', 'child', 'coffe_maker', 'tray', 'comb', 'dresser', 'cupboard', 'hairbrush', 'soap', 'tooth_paste']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet guests when they knock on the door and invite them to enter my home", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sheets', 'tablelamp', 'toilet', 'chair', 'walllamp', 'nail_polish', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'ceilinglamp', 'mat', 'tvstand', 'stereo', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'novel', 'coffe_maker', 'check', 'television', 'curtain', 'lighter', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'beer', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'mat', 'toaster', 'shower', 'nightstand', 'wall', 'powersocket', 'oven', 'cup', 'food_vegetable', 'photoframe', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'tooth_paste']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I prepare meals and cook them on the stove.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[FIND] <oven>", "[SWITCHON] <oven>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'knifeblock', 'knife', 'orchid', 'floor', 'ceiling', 'phone', 'keyboard', 'mail', 'stovefan', 'ceilinglamp', 'deck_of_cards', 'stereo', 'pillow', 'dining_room', 'wall', 'cup', 'bed', 'novel', 'coffe_maker', 'doorjamb', 'kitchen_counter', 'cupboard', 'controller', 'laundry_detergent', 'hairdryer', 'sink', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Making coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'food_carrot', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'television', 'oven_mitts', 'wooden_spoon', 'doorjamb', 'mouse', 'pajamas', 'door', 'light', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Compose an email in the home office.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'toilet', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'toaster', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'love_seat', 'clothes_gloves', 'television', 'tray', 'curtain', 'food_onion', 'dresser', 'cupboard', 'instrument_guitar', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I'll watch TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sheets', 'tablelamp', 'toilet', 'chair', 'walllamp', 'nail_polish', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'ceilinglamp', 'mat', 'tvstand', 'stereo', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'novel', 'coffe_maker', 'check', 'television', 'curtain', 'lighter', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'keyboard', 'ceilinglamp', 'mat', 'picture', 'window', 'couch', 'dining_room', 'wall', 'oven', 'powersocket', 'toy', 'photoframe', 'television', 'oven_mitts', 'tray', 'centerpiece', 'kettle', 'towel_rack', 'pencil', 'dishrack', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room and sit on the couch, ease into the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bathroom', 'detergent', 'food_butter', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'bedroom', 'button', 'table_cloth', 'stovefan', 'mat', 'brush', 'stereo', 'picture', 'mousepad', 'couch', 'clothes_pants', 'food_jam', 'powersocket', 'oven', 'newspaper', 'cards', 'shoes', 'controller', 'clothes_skirt', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into the living room and turn on the radio.  Then, I select the station I want, which is usually a preset.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'video_game_console', 'desk', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'video_game_controller', 'television', 'food_rice', 'ice', 'curtain', 'doorjamb', 'dresser', 'door', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on game.  Use controller,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[PLUGIN] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'glue', 'toilet', 'chair', 'detergent', 'dough', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'keyboard', 'alcohol', 'stovefan', 'ceilinglamp', 'couch', 'microphone', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'doorjamb', 'kitchen_counter', 'hairbrush', 'cupboard', 'home_office', 'dishwasher', 'coffee', 'shoe_rack']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[PUTBACK] <coffee> <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'freezer', 'walllamp', 'bathroom', 'dough', 'knife', 'orchid', 'thread', 'computer', 'ceilinglamp', 'deck_of_cards', 'shower', 'clothes_socks', 'food_noodles', 'dustpan', 'nightstand', 'photoframe', 'light_bulb', 'television', 'coffee_filter', 'mirror', 'curtain', 'doorjamb', 'laundry_detergent', 'sink', 'bag', 'table', 'soap', 'clothes_shirt', 'creditcard']\nTask:Look at mirror\nresponse:\n{\n    \"thoughts\": I look at my reflection to check my cleanliness.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <mirror>', '[FIND] <mirror>', '[TURNTO] <mirror>', '[LOOKAT] <mirror>']\n}\n", "input": "\nScene_Objects:['microwave', 'dog', 'walllamp', 'cat', 'knifeblock', 'drawing', 'bills', 'floor', 'ceiling', 'clothes_jacket', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'iron', 'dish_soap', 'food_noodles', 'mousepad', 'couch', 'wall', 'nightstand', 'mouthwash', 'curtain', 'mouse', 'kitchen_counter', 'cupboard', 'conditioner', 'door', 'sink', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walk to Kitchen, find sink, find dishes and dish soap. Put soap on plate. open faucet, rinse dishes close faucet.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <plate>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'keys', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I want to relax on the sofa to some relaxing music in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <stereo>', '[PLUGIN] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'toilet', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'ceilinglamp', 'mat', 'shower', 'food_noodles', 'couch', 'towel', 'wall', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'newspaper', 'water', 'food_donut', 'curtain', 'doorjamb', 'headset', 'door', 'bathroom_cabinet', 'light', 'soap', 'tooth_paste']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[TURNTO] <shower>", "[FIND] <water>", "[RINSE] <water>", "[FIND] <soap>", "[WASH] <soap>", "[RINSE] <water>", "[FIND] <towel>", "[GRAB] <towel>", "[SCRUB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'water_glass', 'tablelamp', 'microwave', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'tea_bag', 'toaster', 'stovefan', 'mat', 'shower', 'faucet', 'ceilinglamp', 'oil', 'wall', 'oven', 'powersocket', 'coffe_maker', 'food_food', 'water', 'ice', 'tray', 'doorjamb', 'towel_rack', 'coffee_pot', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bag', 'table', 'dishwasher', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Going to the kitchen to get a drink of water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[FIND] <sink>', '[PUTBACK] <water_glass> <sink>', '[GRAB] <water_glass>', '[SWITCHOFF] <faucet>', '[DRINK] <water_glass>', '[FIND] <kitchen_counter>', '[PUTBACK] <water_glass> <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'toilet', 'chair', 'freezer', 'detergent', 'food_butter', 'wallshelf', 'hanger', 'phone', 'character', 'desk', 'remote_control', 'button', 'piano_bench', 'toaster', 'mat', 'tvstand', 'picture', 'oil', 'couch', 'clothes_pants', 'pillow', 'towel', 'wall', 'novel', 'razor', 'diary', 'conditioner', 'table', 'coffee']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'keyboard', 'thread', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'mousepad', 'couch', 'wall', 'photoframe', 'bed', 'toilet_paper', 'food_onion', 'towel_rack', 'mouse', 'kitchen_counter', 'cd', 'hairbrush', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'food_salt', 'microwave', 'freezer', 'orchid', 'drawing', 'hanger', 'button', 'table_cloth', 'tvstand', 'homework', 'couch', 'dining_room', 'nightstand', 'keys', 'coffe_maker', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'diary', 'cards', 'kitchen_counter', 'comb', 'dresser', 'food_steak', 'home_office', 'light', 'clothes_skirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go to the living room then go to the cat and touch it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'newspaper', 'television', 'curtain', 'doorjamb', 'door', 'home_office', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room. look at remote control. grab remote control. plug in television. switch on television. lie in sofa. point at remote control. watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[LIE] <couch>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'spectacles', 'pot', 'microwave', 'knifeblock', 'floor', 'ceiling', 'folder', 'character', 'toaster', 'plate', 'drying_rack', 'faucet', 'dish_soap', 'couch', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'hairdryer', 'light', 'dirt', 'coffee']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I walk in to the Kitchen. I find a plate, a sponge, and the dish soap. I find the faucet. I turn the faucet on. I pour water on the sponge and plate. I add dish soap to the sponge. I put the dish soap back. I scrub the plate. I rinse the plate. I place the plate on the drying rack. I rinse the sponge. I turn the faucet off.", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <sponge>", "[GRAB] <sponge>", "[RINSE] <sponge>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>", "[RINSE] <plate>", "[FIND] <drying_rack>", "[PUTBACK] <plate> <plate>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'bathroom', 'detergent', 'food_butter', 'drawing', 'wallshelf', 'floor', 'desk', 'bookshelf', 'plate', 'faucet', 'picture', 'iron', 'window', 'dish_soap', 'food_jam', 'wall', 'nightstand', 'toy', 'cutting_board', 'food_food', 'television', 'curtain', 'pajamas', 'conditioner', 'light', 'dishwasher', 'tooth_paste', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sponge', 'sheets', 'filing_cabinet', 'crayon', 'toilet', 'bathroom_counter', 'ceiling', 'folder', 'shower', 'iron', 'window', 'mousepad', 'food_peanut_butter', 'pillow', 'wall', 'oven', 'toy', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'food_food', 'tray', 'cd', 'dishrack', 'door', 'napkin', 'light', 'dishwasher', 'board_game']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up your book and read it", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'ceilinglamp', 'laptop', 'faucet', 'mat', 'closetdrawer', 'teeth', 'dining_room', 'wall', 'pasta', 'toy', 'shelf', 'standingmirror', 'novel', 'check', 'television', 'curtain', 'console', 'dresser', 'napkin', 'table', 'light', 'scrabble', 'coffee']\nTask:Look at painting\nresponse:\n{\n    \"thoughts\": Walk to the painting.  Look up at the painting.  Admire it on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'dog', 'toothbrush', 'knifeblock', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'character', 'desk', 'mat', 'tvstand', 'shower', 'homework', 'dining_room', 'wall', 'bed', 'clothes_gloves', 'toothbrush_holder', 'after_shave', 'doorjamb', 'mouse', 'pencil', 'cd', 'tooth_paste', 'shoe_rack']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'walllamp', 'needle', 'wallshelf', 'cpuscreen', 'hanger', 'bedroom', 'keyboard', 'bookmark', 'stovefan', 'shower', 'oil', 'dining_room', 'wall', 'powersocket', 'bench', 'love_seat', 'standingmirror', 'novel', 'coffe_maker', 'check', 'after_shave', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'maindoor', 'bathtub', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'juice', 'food_salt', 'measuring_cup', 'chair', 'detergent', 'milk', 'drawing', 'kitchen_cabinet', 'floor', 'shampoo', 'dvd_player', 'keyboard', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'plate', 'tvstand', 'candle', 'towel', 'food_cheese', 'light_bulb', 'television', 'razor', 'curtain', 'kitchen_counter', 'headset', 'table', 'soap']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Grab plates and forks. Walk into dining room. Put plates and forks on table.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <fork> <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_pizza', 'tablelamp', 'food_ice_cream', 'crayon', 'shampoo', 'keyboard', 'mail', 'bookmark', 'stovefan', 'faucet', 'pillow', 'dining_room', 'broom', 'wall', 'love_seat', 'photoframe', 'novel', 'food_food', 'wine', 'tray', 'after_shave', 'shaving_cream', 'kettle', 'headset', 'pajamas', 'laundry_detergent', 'door', 'bathroom_cabinet', 'shoe_rack']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['food_salt', 'ceiling', 'phone', 'bookshelf', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'homework', 'mousepad', 'food_jam', 'form', 'pillow', 'dining_room', 'wall', 'powersocket', 'bed', 'television', 'tray', 'curtain', 'doorjamb', 'pencil', 'dresser', 'controller', 'laundry_detergent', 'door', 'napkin', 'light', 'dishwasher', 'table', 'teeth']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office. I find the telephone. I pick up the telephone handset.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'piano_bench', 'mat', 'shoe-shine_kit', 'balanceball', 'oil', 'clothes_hat', 'dining_room', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'mouthwash', 'doorjamb', 'towel_rack', 'vacuum_cleaner', 'door', 'home_office', 'light', 'soap']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on the light,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'toilet', 'freezer', 'ground_coffee', 'bathroom', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'bookshelf', 'alcohol', 'mat', 'shower', 'window', 'dining_room', 'food_kiwi', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'television', 'doorjamb', 'mouse', 'door', 'light', 'colander', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I get my groceries and go to the fridge. I open the fridge and put the groceries there.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'character', 'desk', 'tea_bag', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'picture', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'cutting_board', 'novel', 'check', 'television', 'curtain', 'conditioner', 'hairdryer', 'clothes_dress', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": watching TV from sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[DROP] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'mop', 'microwave', 'walllamp', 'detergent', 'food_carrot', 'knifeblock', 'wallshelf', 'hanger', 'floor', 'ceiling', 'keyboard', 'stovefan', 'toaster', 'stereo', 'window', 'oil', 'clothes_socks', 'wall', 'oven', 'food_vegetable', 'bed', 'photoframe', 'television', 'food_donut', 'water', 'tray', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "walk to kicthen , find stove , find cooking pot , find vegetables , find water  ,put vegetables in cooking pot, pour water to cooking pot,put cooking pot on stove, switch on stove", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[FIND] <pot>", "[FIND] <food_vegetable>", "[FIND] <water>", "[GRAB] <water>", "[WASH] <food_vegetable>", "[POUR] <water> <pot>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'cat', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'mousepad', 'bench', 'newspaper', 'coffe_maker', 'food_food', 'tray', 'curtain', 'mouse', 'door', 'hairdryer', 'sink', 'food_steak', 'light', 'dishwasher', 'spoon', 'table', 'board_game']\nTask:Write  school paper\nresponse:\n{\n    \"thoughts\": Writing  and drafting a paper for  class work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'cat', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'ceilinglamp', 'homework', 'mousepad', 'wall', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'shoes', 'comb', 'dresser', 'hairdryer']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I open my laptop and then login to Windows. Then I open Google Chrome, and type gmail.com into the address bar. My email automatically logs in, then I click the Compose button, type in the address of the recipient, then the subject of the email, and then start typing my email", "actions": ["[WALK] <laptop>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'coffe_maker', 'television', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'home_office', 'table', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the living room and find the book you want to read. Take it and then go to the couch. Sit on it, open the book and start reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'blender', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'floor', 'bedroom', 'bookshelf', 'stovefan', 'food_peanut_butter', 'clothes_hat', 'clothes_pants', 'dining_room', 'nightstand', 'wall_clock', 'video_game_controller', 'food_donut', 'curtain', 'kettle', 'shoes', 'dresser', 'cupboard', 'hairbrush', 'controller', 'home_office', 'light', 'clothes_skirt', 'soap']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the radio. Turn on the radio. Listen to the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'orchid', 'drawing', 'hanger', 'floor', 'phone', 'character', 'keyboard', 'toaster', 'plate', 'faucet', 'shower', 'window', 'mousepad', 'closetdrawer', 'dining_room', 'oven', 'powersocket', 'cup', 'bed', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'dresser', 'sink', 'bathtub', 'home_office', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'bathroom', 'knifeblock', 'cpuscreen', 'hanger', 'bathroom_counter', 'ceiling', 'folder', 'character', 'keyboard', 'bookshelf', 'envelope', 'computer', 'dining_room', 'wall', 'nightstand', 'cup', 'clothes_gloves', 'light_bulb', 'check', 'curtain', 'pencil', 'kitchen_counter', 'controller', 'sink', 'napkin', 'clothes_skirt', 'tooth_paste']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I turn on computer. I use keyboard to type. I open browser. I search.", "actions": ["[WALK] <bedroom>", "[WALK] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'cat', 'drawing', 'knife', 'knifeblock', 'hanger', 'floor', 'bedroom', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'faucet', 'iron', 'window', 'mousepad', 'couch', 'form', 'dining_room', 'wall', 'oven', 'box', 'newspaper', 'bed', 'novel', 'toilet_paper', 'television', 'tray', 'sink', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'measuring_cup', 'microwave', 'dog', 'orchid', 'wallshelf', 'floor', 'bedroom', 'computer', 'mop_bucket', 'ceilinglamp', 'faucet', 'electric_shaver', 'couch', 'food_peanut_butter', 'broom', 'cutting_board', 'photoframe', 'newspaper', 'mouthwash', 'ice', 'tray', 'curtain', 'kitchen_counter', 'shoes', 'headset', 'bathroom_cabinet', 'board_game']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my living room, find my cat, scrub it using my hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'toilet', 'chair', 'bathroom', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'computer', 'tvstand', 'faucet', 'stereo', 'food_orange', 'photoframe', 'novel', 'television', 'curtain', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'shoes', 'dresser', 'cupboard', 'sink', 'bathtub', 'home_office']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": I get dressed in my bedroom in the mornings.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'folder', 'desk', 'alcohol', 'stovefan', 'plate', 'mousepad', 'wall', 'oven', 'food_food', 'check', 'tray', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'sink', 'clothes_scarf', 'light', 'clothes_shirt', 'teeth']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk over and pick up the phone", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'microwave', 'freezer', 'walllamp', 'detergent', 'toothbrush', 'orchid', 'drawing', 'clothes_underwear', 'floor', 'thread', 'mat', 'closetdrawer', 'wall', 'nightstand', 'toy', 'standingmirror', 'light_bulb', 'novel', 'check', 'tray', 'cards', 'mouse', 'pencil', 'vacuum_cleaner', 'headset', 'maindoor', 'napkin', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the Living Room. Look at the Cabinet. Find a Textbook. Grab the textbook. Open the Textbook. Read the textbook.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[FIND] <filing_cabinet>', '[TURNTO] <filing_cabinet>', '[LOOKAT] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'blow_dryer', 'bathroom', 'nail_polish', 'drawing', 'knife', 'cpuscreen', 'character', 'bookshelf', 'table_cloth', 'mop_bucket', 'bookmark', 'faucet', 'homework', 'nightstand', 'wall', 'novel', 'check', 'coffee_filter', 'coffee_pot', 'shaving_cream', 'kitchen_counter', 'pajamas', 'cupboard', 'clothes_dress', 'laundry_detergent', 'napkin', 'table', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "walk to kitchen. grab clothes_dress. look at washing machine. open washing machine. put clothes_dress on washing machine. close washing machine. switch on washing machine. switch off washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <clothes_dress>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[TURNTO] <washing_machine>", "[LOOKAT] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>", "[SWITCHOFF] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'mop', 'walllamp', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'dvd_player', 'desk', 'tvstand', 'window', 'clothes_socks', 'closetdrawer', 'oil', 'dining_room', 'wall', 'toy', 'photoframe', 'television', 'curtain', 'rag', 'doorjamb', 'console', 'kitchen_counter', 'shoes', 'dresser', 'pajamas', 'bathroom_cabinet', 'table', 'tooth_paste']\nTask:Put away shoes\nresponse:\n{\n    \"thoughts\": Pick up shoes from entry hall. Open entry way closet. Put shoes in closet. Close closet door.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[FIND] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'mop', 'bathroom', 'dough', 'knifeblock', 'hanger', 'keyboard', 'alcohol', 'toaster', 'shower', 'foundation', 'couch', 'bowl', 'wall', 'pasta', 'basket_for_clothes', 'check', 'television', 'razor', 'tray', 'doorjamb', 'chessboard', 'dishrack', 'hairbrush', 'napkin', 'bathroom_cabinet', 'light', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "go to living room, find phone, pick up phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'video_game_console', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'dustpan', 'pillow', 'wall', 'cup', 'video_game_controller', 'television', 'curtain', 'dishrack', 'napkin', 'table', 'vase', 'scrabble', 'shoe_rack']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on game.  Use controller,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[PLUGIN] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'food_cake', 'stovefan', 'mat', 'toaster', 'shower', 'food_noodles', 'oil', 'nightstand', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'juice', 'microwave', 'knifeblock', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'piano_bench', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'stovefan', 'tvstand', 'dish_soap', 'wall', 'pasta', 'coffee_filter', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'sink', 'scrabble']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <sink>', '[SWITCHOFF] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[GRAB] <plate>', '[SCRUB] <plate>', '[PUTOBJBACK] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[PUTOBJBACK] <plate>', '[SWITCHON] <faucet>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[GRAB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'chair', 'walllamp', 'blow_dryer', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'ceiling', 'thread', 'ceilinglamp', 'laptop', 'faucet', 'mousepad', 'couch', 'candle', 'food_jam', 'pillow', 'wall', 'nightstand', 'tape', 'television', 'tray', 'after_shave', 'food_onion', 'shoes', 'cupboard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I sit in the sofa and I relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'band-aids', 'walllamp', 'drawing', 'food_egg', 'bathroom_counter', 'ceiling', 'stamp', 'desk', 'cleaning_solution', 'computer', 'stovefan', 'deck_of_cards', 'faucet', 'tvstand', 'window', 'couch', 'wall', 'nightstand', 'coffe_maker', 'television', 'mirror', 'after_shave', 'paper_towel', 'clothes_dress', 'sink', 'bag', 'bathroom_cabinet', 'dishwasher', 'laser_pointer']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[FIND] <cleaning_solution>', '[RINSE] <cleaning_solution>', '[WIPE] <mirror>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'chair', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'bedroom', 'button', 'plate', 'dish_soap', 'wall', 'nightstand', 'food_cheese', 'photoframe', 'newspaper', 'coffe_maker', 'check', 'tray', 'kettle', 'pencil', 'kitchen_counter', 'comb', 'cupboard', 'door', 'table', 'dishwasher', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'spectacles', 'microwave', 'walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'character', 'desk', 'alcohol', 'stovefan', 'ceilinglamp', 'mat', 'window', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'cutting_board', 'photoframe', 'light_bulb', 'television', 'curtain', 'kettle', 'napkin', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_salt', 'microwave', 'walllamp', 'dough', 'hanger', 'ceiling', 'phone', 'character', 'keyboard', 'bookmark', 'toaster', 'mat', 'tvstand', 'stereo', 'window', 'oil', 'food_peanut_butter', 'tape', 'check', 'video_game_controller', 'razor', 'oven_mitts', 'food_rice', 'curtain', 'kitchen_counter', 'hairbrush', 'food_steak', 'bathroom_cabinet', 'home_office']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the bedroom. I walk to stereo. I turn on stereo.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'stereo', 'shower', 'mousepad', 'closetdrawer', 'dustpan', 'wall', 'powersocket', 'box', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'light', 'table', 'printing_paper']\nTask:Print out papers\nresponse:\n{\n    \"thoughts\": Turn on printer and fill with blank paper. Use computer to click on printing option. Take printed page out of print.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <fax_machine>', '[FIND] <fax_machine>', '[SWITCHON] <fax_machine>', '[FIND] <printing_paper>', '[GRAB] <printing_paper>', '[PUTBACK] <printing_paper> <fax_machine>', '[FIND] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[GRAB] <printing_paper>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'blender', 'beer', 'toilet', 'jelly', 'freezer', 'blow_dryer', 'detergent', 'orchid', 'knifeblock', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'mail', 'computer', 'bookmark', 'stereo', 'window', 'couch', 'microphone', 'check', 'television', 'rag', 'pencil', 'controller', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Pet Cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'bathroom', 'knife', 'bathroom_counter', 'floor', 'shampoo', 'ceiling', 'stovefan', 'mat', 'faucet', 'mousepad', 'couch', 'food_peanut_butter', 'dustpan', 'towel', 'wall', 'wall_clock', 'oven', 'novel', 'ice', 'tray', 'rag', 'shoes', 'maindoor', 'sink', 'bathroom_cabinet', 'vase', 'soap', 'tooth_paste', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'walllamp', 'detergent', 'orchid', 'floor', 'ceiling', 'keyboard', 'remote_control', 'computer', 'button', 'alcohol', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'check', 'mouthwash', 'television', 'coffe_maker', 'curtain', 'dresser', 'clothes_dress', 'door', 'sink', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'chair', 'walllamp', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'bedroom', 'desk', 'computer', 'ceilinglamp', 'deck_of_cards', 'stereo', 'window', 'couch', 'clothes_pants', 'wall', 'cup', 'novel', 'toilet_paper', 'television', 'wine', 'kitchen_counter', 'dishrack', 'comb', 'dresser', 'coffee_cup', 'table', 'teeth']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'crayon', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'food_cake', 'ceilinglamp', 'tvstand', 'mat', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'food_food', 'television', 'curtain', 'pencil', 'kitchen_counter', 'door']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the sofa, and lie down on the sofa.", "actions": ["[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'mop', 'toilet', 'cat', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'orchid', 'floor', 'ceiling', 'computer', 'clothes_jacket', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'couch', 'wall', 'nightstand', 'food_food', 'mouthwash', 'television', 'food_donut', 'curtain', 'after_shave', 'dresser', 'table']\nTask:Look at painting\nresponse:\n{\n    \"thoughts\": Walk to the painting.  Look up at the painting.  Admire it on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'toilet', 'freezer', 'walllamp', 'detergent', 'milk', 'drawing', 'floor', 'ceiling', 'phone', 'computer', 'toaster', 'mat', 'ceilinglamp', 'shower', 'food_orange', 'pillow', 'dining_room', 'nightstand', 'wall', 'powersocket', 'photoframe', 'newspaper', 'food_food', 'toothbrush_holder', 'doorjamb', 'door', 'sink', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk to the kitchen, open the fridge and put the groceries in the fridge. Then I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_ice_cream', 'walllamp', 'nail_polish', 'cat', 'orchid', 'drawing', 'floor', 'ceiling', 'purse', 'alcohol', 'mail', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'towel', 'wall', 'oven', 'cup', 'food_cheese', 'television', 'curtain', 'doorjamb', 'door', 'table', 'clothes_shirt']\nTask:Straighten paintings on wall\nresponse:\n{\n    \"thoughts\": Go up to the painting and tilt it which ever way it is uneven until it looks straight.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <wall>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'orchid', 'knifeblock', 'wallshelf', 'character', 'keyboard', 'stovefan', 'mat', 'closetdrawer', 'dining_room', 'food_kiwi', 'box', 'cup', 'bed', 'coffe_maker', 'television', 'water', 'colander', 'tray', 'curtain', 'kitchen_counter', 'cd', 'cupboard', 'door', 'bag', 'soap', 'creditcard']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "put soap on hands, scrub it and rinse with water", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'drawing', 'bathroom_counter', 'bedroom', 'keyboard', 'button', 'toaster', 'tvstand', 'food_noodles', 'pillow', 'check', 'television', 'oven_mitts', 'wine', 'curtain', 'rag', 'doorjamb', 'kettle', 'mouse', 'pencil', 'hairbrush', 'food_steak', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find the cat; pet the cat by touching it,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_bread', 'microwave', 'toilet', 'walllamp', 'knifeblock', 'wallshelf', 'floor', 'shampoo', 'character', 'computer', 'bookmark', 'toaster', 'plate', 'faucet', 'shoe-shine_kit', 'dish_soap', 'wall', 'toy', 'light_bulb', 'coffe_maker', 'rag', 'kitchen_counter', 'hairbrush', 'cupboard', 'dresser', 'controller', 'sink', 'home_office', 'table', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Make some dish water, grab a dish rag, put dishes in dish water and wash.", "actions": ["[RUN] <sink>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[PUTBACK] <dish_soap> <plate>", "[RINSE] <dish_soap>", "[FIND] <plate>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'keyboard', 'mail', 'mat', 'faucet', 'wall', 'nightstand', 'powersocket', 'oven', 'cup', 'photoframe', 'bed', 'novel', 'curtain', 'doorjamb', 'diary', 'cupboard', 'door', 'sink', 'instrument_guitar']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk over to the bookshelf. Pick storybook off the shelf. Sit on bed. Open storybook and read out loud.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[WALK] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'food_ice_cream', 'jelly', 'bathroom', 'orchid', 'knifeblock', 'wallshelf', 'bathroom_counter', 'desk', 'keyboard', 'ceilinglamp', 'iron', 'stereo', 'closetdrawer', 'clothes_pants', 'food_jam', 'cup', 'photoframe', 'toothbrush_holder', 'tray', 'curtain', 'centerpiece', 'coffee_pot', 'kitchen_counter', 'door', 'spoon', 'dishwasher', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter living room. Walk to telephone. Pick up from hook.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'chair', 'jelly', 'walllamp', 'drawing', 'wallshelf', 'food_egg', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'food_rice', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'coffee']\nTask:Watch  horror  movie\nresponse:\n{\n    \"thoughts\": I would  find something  scary  to  watch on dvd,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'water_glass', 'pot', 'glue', 'filing_cabinet', 'knife', 'wallshelf', 'bathroom_counter', 'stamp', 'bedroom', 'keyboard', 'clothes_jacket', 'piano_bench', 'electric_shaver', 'food_jam', 'cutting_board', 'novel', 'check', 'coffee_filter', 'water', 'ice', 'curtain', 'food_onion', 'pencil', 'kitchen_counter', 'cd', 'headset', 'dresser', 'door', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drinking a glass of water", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'food_bread', 'tablelamp', 'scissors', 'microwave', 'chair', 'walllamp', 'hanger', 'floor', 'food_cake', 'toaster', 'plate', 'shower', 'tvstand', 'faucet', 'dish_soap', 'bowl', 'wall', 'oven', 'cup', 'basket_for_clothes', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'curtain', 'maindoor', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Using dishwasher to wash dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dishwasher>', '[PUTBACK] <bowl> <dishwasher>', '[GRAB] <bowl>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'blender', 'food_salt', 'crayon', 'tea', 'cat', 'milk', 'knifeblock', 'hanger', 'shampoo', 'bedroom', 'bookshelf', 'computer', 'brush', 'stovefan', 'toaster', 'foundation', 'window', 'mousepad', 'clothes_pants', 'clothes_hat', 'tape', 'bed', 'coffe_maker', 'food_onion', 'after_shave', 'dresser', 'light', 'dishwasher', 'laser_pointer']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Every time this animal is seen, you must approach it and show compassion for it by petting it in a gentle manner.", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <bedroom>", "[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[FIND] <brush>", "[GRAB] <brush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'keyboard', 'mat', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'cupboard', 'door', 'light', 'table']\nTask:Gaze out window\nresponse:\n{\n    \"thoughts\": Look at the yard from the window,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <window>', '[FIND] <window>', '[TURNTO] <window>', '[LOOKAT] <window>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'food_pizza', 'knifeblock', 'food_egg', 'hanger', 'phone', 'desk', 'tea_bag', 'purse', 'envelope', 'clothes_jacket', 'table_cloth', 'faucet', 'window', 'closetdrawer', 'clothes_pants', 'food_jam', 'wall', 'basket_for_clothes', 'bed', 'cd', 'comb', 'laundry_detergent', 'clothes_scarf', 'napkin', 'home_office', 'bathroom_cabinet', 'washing_machine', 'teeth']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[PLUGIN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'food_bread', 'tablelamp', 'scissors', 'microwave', 'chair', 'walllamp', 'hanger', 'floor', 'food_cake', 'toaster', 'plate', 'shower', 'tvstand', 'faucet', 'dish_soap', 'bowl', 'wall', 'oven', 'cup', 'basket_for_clothes', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'curtain', 'maindoor', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Using dishwasher to wash dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dishwasher>', '[PUTBACK] <bowl> <dishwasher>', '[GRAB] <bowl>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'dog', 'ground_coffee', 'detergent', 'floor', 'ceiling', 'shampoo', 'envelope', 'clothes_jacket', 'piano_bench', 'ceilinglamp', 'mat', 'couch', 'dining_room', 'wall', 'coffe_maker', 'food_food', 'television', 'coffee_filter', 'water', 'coffee_pot', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'pajamas', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making Coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <water>", "[GRAB] <water>", "[OPEN] <coffe_maker>", "[POUR] <water> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <coffee_pot>", "[GRAB] <coffee_pot>", "[PUTBACK] <coffee_pot> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_ice_cream', 'beer', 'crayon', 'chair', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'phone', 'desk', 'bedroom', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'cup', 'bed', 'keys', 'food_food', 'television', 'food_onion', 'doorjamb', 'mouse', 'home_office']\nTask:Lock door\nresponse:\n{\n    \"thoughts\": Turn the lock.,\n    \"actions\": ['[WALK] <bedroom>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'toilet', 'chair', 'walllamp', 'floor', 'ceiling', 'phone', 'keyboard', 'envelope', 'toaster', 'ceilinglamp', 'stereo', 'faucet', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'clothes_gloves', 'novel', 'check', 'video_game_controller', 'television', 'kitchen_counter', 'hairdryer', 'light', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Walk to the sofa and sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'chair', 'walllamp', 'blow_dryer', 'knifeblock', 'drawing', 'orchid', 'kitchen_cabinet', 'floor', 'ceiling', 'food_cake', 'mat', 'faucet', 'mousepad', 'couch', 'wall', 'oven', 'bench', 'toilet_paper', 'television', 'water', 'drinking_glass', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen to get a glass from the cabinet. I turn on the faucet and fill up the glass. I take a drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <drinking_glass> <sink>', '[FIND] <water>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'milk', 'orchid', 'drawing', 'knifeblock', 'floor', 'ceiling', 'folder', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'cup', 'bed', 'television', 'razor', 'toothbrush_holder', 'tray', 'curtain', 'dresser', 'laundry_detergent', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to the living room to watch TV", "actions": ["[RUN] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[TOUCH] <remote_control>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[DROP] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_pizza', 'walllamp', 'orchid', 'needle', 'cpuscreen', 'bathroom_counter', 'shampoo', 'desk', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'shower', 'faucet', 'couch', 'dining_room', 'cup', 'bed', 'tray', 'curtain', 'wooden_spoon', 'pencil', 'door', 'bathroom_cabinet', 'light', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office. I find the telephone. I pick up the telephone handset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'oil', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'nightstand', 'wall', 'oven', 'bed', 'coffe_maker', 'television', 'curtain', 'dresser', 'door', 'instrument_guitar', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I go to the living room, switch on the tv, take the remote and sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'laptop', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'tape', 'photoframe', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'light']\nTask:Use laptop\nresponse:\n{\n    \"thoughts\": Sit down on a couch and perform tasks such as navigating the internet on a laptop,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <keyboard>', '[LOOKAT] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['chair', 'detergent', 'food_butter', 'clothes_underwear', 'hanger', 'floor', 'stamp', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'shoe-shine_kit', 'clothes_socks', 'closetdrawer', 'couch', 'clothes_pants', 'basket_for_clothes', 'check', 'mouthwash', 'food_rice', 'vacuum_cleaner', 'laundry_detergent', 'clothes_dress', 'bathroom_cabinet', 'light', 'dishwasher', 'clothes_shirt', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[DROP] <basket_for_clothes>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[GRAB] <laundry_detergent>", "[PUTOBJBACK] <laundry_detergent>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'mousepad', 'couch', 'towel', 'wall', 'bed', 'photoframe', 'curtain', 'after_shave', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'light', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap and hot water,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['juice', 'microwave', 'toilet', 'ground_coffee', 'bathroom', 'cpuscreen', 'hanger', 'floor', 'phone', 'character', 'bookshelf', 'computer', 'table_cloth', 'mat', 'tvstand', 'shower', 'couch', 'pillow', 'bowl', 'wall', 'powersocket', 'novel', 'food_food', 'video_game_controller', 'check', 'food_onion', 'doorjamb', 'napkin', 'home_office', 'coffee']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the living room. I find the stereo and turn it on, and I listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'alcohol', 'toaster', 'tvstand', 'shower', 'closetdrawer', 'nightstand', 'wall_clock', 'bench', 'powersocket', 'food_food', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'sink', 'home_office', 'table', 'light', 'coffee']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the hallway.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dining_room>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'microwave', 'walllamp', 'toothbrush', 'drawing', 'knifeblock', 'wallshelf', 'keyboard', 'food_cake', 'ceilinglamp', 'tvstand', 'shower', 'stereo', 'oil', 'tooth_paste', 'couch', 'closetdrawer', 'dining_room', 'toy', 'basket_for_clothes', 'light_bulb', 'check', 'centerpiece', 'kettle', 'mouse', 'kitchen_counter', 'shoes', 'light', 'board_game']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "EARLY MORNING AND BEFORE GOING TO BED", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'tea', 'beer', 'microwave', 'toilet', 'trashcan', 'food_carrot', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'bedroom', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'window', 'dining_room', 'food_kiwi', 'nightstand', 'bed', 'photoframe', 'doorjamb', 'dishrack', 'cupboard', 'sink', 'home_office']\nTask:Open bathroom window\nresponse:\n{\n    \"thoughts\": Make sure bathroom window is unlocked. Lift up bathroom window so it is open.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <window>', '[FIND] <window>', '[OPEN] <window>']\n}\n", "input": "\nScene_Objects:['blender', 'toilet', 'freezer', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'shoe-shine_kit', 'shower', 'faucet', 'mat', 'oil', 'food_peanut_butter', 'towel', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'toilet_paper', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'sink', 'bathroom_cabinet', 'light', 'colander']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <towel>", "[GRAB] <towel>", "[CLOSE] <bathroom_cabinet>", "[FIND] <shower>", "[RINSE] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'toilet', 'freezer', 'band-aids', 'knifeblock', 'food_egg', 'shampoo', 'phone', 'bookshelf', 'purse', 'thread', 'deck_of_cards', 'shower', 'picture', 'clothes_socks', 'oven', 'food_cheese', 'novel', 'food_food', 'check', 'coffe_maker', 'tray', 'after_shave', 'curtain', 'doorjamb', 'cupboard', 'controller', 'bag', 'table', 'scrabble']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Pour water into a glass and drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'walllamp', 'drawing', 'needle', 'orchid', 'floor', 'stamp', 'keyboard', 'plate', 'faucet', 'shower', 'dish_soap', 'oil', 'wall', 'keys', 'coffe_maker', 'coffee_filter', 'pencil', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'controller', 'pajamas', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'toilet', 'freezer', 'dough', 'drawing', 'hanger', 'floor', 'ceiling', 'mat', 'shoe-shine_kit', 'shower', 'clothes_hat', 'bowl', 'nightstand', 'wall', 'powersocket', 'wall_clock', 'photoframe', 'light_bulb', 'food_food', 'coffe_maker', 'razor', 'coffee_filter', 'doorjamb', 'hairbrush', 'door', 'home_office', 'light', 'soap']\nTask:Put away groceries\nresponse:\n{\n    \"thoughts\": Place groceries in refrigerator from shopping at the grocery store.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['beer', 'walllamp', 'blow_dryer', 'knifeblock', 'hanger', 'floor', 'ceiling', 'piano_bench', 'ceilinglamp', 'faucet', 'stereo', 'oil', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'coffe_maker', 'toilet_paper', 'coffee_filter', 'water', 'wine', 'tray', 'curtain', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'dishwasher', 'laser_pointer', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Go to the kitchen and make coffee.", "actions": ["[RUN] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[PUTOBJBACK] <coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'plate', 'stereo', 'mousepad', 'wall', 'bench', 'television', 'wine', 'rag', 'lighter', 'shaving_cream', 'mouse', 'hairbrush', 'door', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": walk to living room, look at computer, find chair, sit in chair, grab keyboard, grab mouse, type using keyboard and mouse,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['beer', 'toilet', 'freezer', 'bathroom', 'knifeblock', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'clothes_jacket', 'mat', 'shower', 'candle', 'creditcard', 'food_jam', 'food_orange', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_rice', 'doorjamb', 'door', 'spoon', 'light', 'board_game', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Grab groceries and walk to the fridge. Open fridge and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'cat', 'drawing', 'hanger', 'longboard', 'floor', 'phone', 'character', 'desk', 'purse', 'stovefan', 'ceilinglamp', 'iron', 'stereo', 'window', 'oil', 'couch', 'wall', 'bench', 'toy', 'food_food', 'television', 'food_rice', 'towel_rack', 'console', 'hairbrush', 'napkin', 'table', 'vase', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I would walk into the living room. Then sit on the couch and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sponge', 'fork', 'microwave', 'toilet', 'walllamp', 'detergent', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bedroom', 'keyboard', 'table_cloth', 'toaster', 'laptop', 'plate', 'faucet', 'stereo', 'shower', 'food_noodles', 'wall', 'nightstand', 'novel', 'check', 'food_food', 'coin', 'kitchen_counter', 'cupboard', 'sink']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walk into the kitchen. Walk up to the sink. Turn on the faucet. Pick up the plate. Wash the plate. Rinse the plate.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>", "[RINSE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'freezer', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'character', 'bedroom', 'keyboard', 'computer', 'stereo', 'shower', 'towel', 'wall', 'wall_clock', 'pasta', 'nightstand', 'bed', 'doorjamb', 'dresser', 'cupboard', 'conditioner', 'door', 'clothes_scarf', 'light', 'clothes_shirt']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": approach the light switch, raise hand to the switch, flip the switch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'microwave', 'freezer', 'walllamp', 'blow_dryer', 'detergent', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'faucet', 'mousepad', 'wall', 'wall_clock', 'oven', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'cards', 'kitchen_counter', 'cupboard', 'sink', 'light', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, grab a cup, fill with water and then drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[FIND] <sink>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'tea', 'mop', 'microwave', 'toilet', 'toothbrush', 'drawing', 'knifeblock', 'orchid', 'phone', 'ceiling', 'computer', 'stovefan', 'toaster', 'tvstand', 'window', 'floor_lamp', 'couch', 'clothes_hat', 'photoframe', 'after_shave', 'food_onion', 'headset', 'chessboard', 'cupboard', 'controller', 'door', 'bathroom_cabinet', 'clothes_skirt']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_salt', 'beer', 'chair', 'bathroom', 'cpuscreen', 'hanger', 'ceiling', 'desk', 'ceilinglamp', 'mat', 'shower', 'stereo', 'window', 'oil', 'dustpan', 'microphone', 'dining_room', 'novel', 'check', 'food_food', 'razor', 'ice', 'curtain', 'rag', 'after_shave', 'coffee_pot', 'shoes', 'instrument_guitar', 'tooth_paste']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I find a radio. I turn on the radio. I listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'toilet', 'chair', 'detergent', 'bathroom_counter', 'ceiling', 'character', 'desk', 'laptop', 'tvstand', 'faucet', 'electric_shaver', 'mousepad', 'couch', 'food_orange', 'bench', 'powersocket', 'cup', 'bed', 'novel', 'razor', 'oven_mitts', 'curtain', 'after_shave', 'cd', 'headset', 'dresser', 'pajamas', 'bathroom_cabinet', 'tooth_paste']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[WALK] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'chair', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'tape', 'photoframe', 'coffee_table', 'television', 'toothbrush_holder', 'razor', 'curtain', 'dresser', 'pajamas', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <coffee_table>", "[PUTBACK] <remote_control> <coffee_table>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'bathroom', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'knifeblock', 'floor', 'dvd_player', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'faucet', 'electric_shaver', 'floor_lamp', 'mousepad', 'wall', 'powersocket', 'novel', 'food_food', 'check', 'television', 'after_shave', 'cupboard', 'sink', 'home_office', 'table', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Twist the lamp switch until the light goes on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'ground_coffee', 'bathroom', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'bookshelf', 'button', 'mat', 'shower', 'mousepad', 'oil', 'closetdrawer', 'microphone', 'oven', 'cutting_board', 'check', 'centerpiece', 'diary', 'clothes_dress', 'clothes_scarf', 'napkin', 'tooth_paste']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the cat. Pet the cat with your hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'door', 'table', 'light', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go in the living room. Then I find the remote, sit on the couch and turn on the television. Then I use the remote the change the TV channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'walllamp', 'food_carrot', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'button', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'basket_for_clothes', 'food_food', 'television', 'curtain', 'dresser', 'clothes_dress', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Take TV remote to change TV Channel. Press channel/program +/- button on TV remote.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[TOUCH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'button', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'faucet', 'mousepad', 'dining_room', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'rag', 'doorjamb', 'kitchen_counter', 'light', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_salt', 'chair', 'ground_coffee', 'bathroom', 'floor', 'ceiling', 'button', 'mop_bucket', 'homework', 'window', 'clothes_hat', 'clothes_pants', 'wall', 'toy', 'novel', 'coffe_maker', 'food_food', 'razor', 'wooden_spoon', 'doorjamb', 'diary', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'coffee_cup', 'dishwasher', 'clothes_skirt', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to the kitchen at the coffee maker, grab coffee, pour coffee into coffee maker, put coffee into coffee maker, push coffee maker's button, wait to complete, grab coffee cup", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[OPEN] <coffe_maker>", "[POUR] <coffee> <coffe_maker>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'chair', 'nail_polish', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'ceilinglamp', 'balanceball', 'shower', 'mousepad', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'toothbrush_holder', 'doorjamb', 'mouse', 'light', 'vase']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['food_bread', 'chair', 'freezer', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'stereo', 'faucet', 'oil', 'mousepad', 'couch', 'closetdrawer', 'dining_room', 'wall', 'oven', 'photoframe', 'coffee_table', 'novel', 'check', 'tray', 'food_onion', 'diary', 'mouse', 'cupboard', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I walk to the living room and pick up a book from the coffee table. I find my favorite chair.  I sit in the armchair and open the book.  I read.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[WALK] <coffee_table>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'chair', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'table_cloth', 'toaster', 'stovefan', 'tvstand', 'shower', 'faucet', 'closetdrawer', 'nightstand', 'wall', 'bench', 'coffe_maker', 'check', 'television', 'tray', 'pencil', 'shoes', 'sink', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": turn radio on, listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tea', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'desk', 'keyboard', 'computer', 'toaster', 'laptop', 'mat', 'stereo', 'mousepad', 'couch', 'wall', 'oven', 'bed', 'centerpiece', 'doorjamb', 'mouse', 'cd', 'dresser', 'food_steak', 'board_game', 'shoe_rack']\nTask:Work\nresponse:\n", "output": {"thoughts": "first ill go to my home office adn ill switch on my laptop and with the use of mouse and keyboard il run my computer", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[RUN] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'pot', 'food_salt', 'mop', 'freezer', 'walllamp', 'detergent', 'knifeblock', 'floor', 'ceiling', 'purse', 'button', 'piano_bench', 'food_bacon', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'form', 'wall', 'bench', 'food_food', 'television', 'water', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'maindoor', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_ice_cream', 'toilet', 'freezer', 'bathroom', 'cpuscreen', 'bathroom_counter', 'phone', 'character', 'desk', 'thread', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'washing_machine', 'window', 'mousepad', 'clothes_pants', 'oven', 'bed', 'keys', 'check', 'dishrack', 'comb', 'laundry_detergent', 'door', 'napkin', 'clothes_shirt', 'dirt', 'coffee']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk into laundry room. Open washing machine door. Put dirty clothes into washing machine. Measure laundry detergent and place into washing machine. Turn dial to appropriate wash setting. Power on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[CLOSE] <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[OPEN] <washing_machine>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'walllamp', 'hanger', 'floor', 'ceiling', 'keyboard', 'computer', 'stovefan', 'stereo', 'shower', 'window', 'mousepad', 'wall', 'nightstand', 'wall_clock', 'toy', 'photoframe', 'coffe_maker', 'curtain', 'lighter', 'towel_rack', 'kitchen_counter', 'controller', 'bathtub', 'bathroom_cabinet', 'table', 'instrument_violin', 'creditcard']\nTask:Open bathroom window\nresponse:\n{\n    \"thoughts\": Make sure bathroom window is unlocked. Lift up bathroom window so it is open.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <window>', '[FIND] <window>', '[OPEN] <window>']\n}\n", "input": "\nScene_Objects:['dog', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'table_cloth', 'mat', 'faucet', 'shower', 'oil', 'clothes_socks', 'dustpan', 'wall', 'toy', 'bed', 'razor', 'water', 'coffee_pot', 'cupboard', 'hairdryer', 'sink', 'napkin', 'bathroom_cabinet', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Wash the teeth", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[TURNTO] <water>", "[POINTAT] <water>", "[FIND] <teeth>", "[WASH] <teeth>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'homework', 'towel', 'dining_room', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bag', 'napkin', 'soap']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen counter, grab a grocery, walk to fridge, open fridge, put the grocery into fridge, and close fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'freezer', 'ground_coffee', 'walllamp', 'food_egg', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'stovefan', 'plate', 'mousepad', 'wall', 'nightstand', 'powersocket', 'novel', 'coffe_maker', 'check', 'coffee_filter', 'water', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'soap', 'teeth', 'shoe_rack']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'detergent', 'cat', 'food_butter', 'knife', 'phone', 'character', 'table_cloth', 'mat', 'deck_of_cards', 'stereo', 'form', 'pillow', 'dining_room', 'nightstand', 'wall', 'basket_for_clothes', 'photoframe', 'tray', 'doorjamb', 'kitchen_counter', 'chessboard', 'dresser', 'clothes_dress', 'table', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_pizza', 'food_salt', 'freezer', 'drawing', 'knife', 'cpuscreen', 'knifeblock', 'floor', 'character', 'bookshelf', 'laptop', 'plate', 'mousepad', 'candle', 'dining_room', 'wall', 'light_bulb', 'food_food', 'coffe_maker', 'tray', 'kitchen_counter', 'chessboard', 'dishrack', 'shoes', 'cupboard', 'laundry_detergent', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'blow_dryer', 'nail_polish', 'ground_coffee', 'milk', 'knife', 'wallshelf', 'hanger', 'character', 'desk', 'keyboard', 'stovefan', 'mat', 'plate', 'food_jam', 'nightstand', 'wall', 'toy', 'bed', 'coffe_maker', 'wine', 'rag', 'centerpiece', 'pencil', 'dresser', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Pick up the book, open to current page, read the words.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'scissors', 'microwave', 'walllamp', 'nail_polish', 'detergent', 'blow_dryer', 'orchid', 'hanger', 'bathroom_counter', 'character', 'desk', 'stamp', 'computer', 'stereo', 'mousepad', 'pillow', 'powersocket', 'toy', 'check', 'food_food', 'television', 'oven_mitts', 'curtain', 'shaving_cream', 'pencil', 'cd', 'cupboard', 'instrument_guitar', 'soap']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find radio, turn it on", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'freezer', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'microphone', 'pillow', 'wall', 'wall_clock', 'nightstand', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'home_office', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room and sit on the couch, ease into the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'crayon', 'chair', 'cat', 'dough', 'milk', 'hanger', 'floor', 'keyboard', 'mail', 'table_cloth', 'stovefan', 'plate', 'faucet', 'tvstand', 'window', 'couch', 'form', 'dining_room', 'cutting_board', 'photoframe', 'check', 'wine', 'dresser', 'hairbrush', 'pajamas', 'conditioner', 'home_office', 'bathroom_cabinet', 'table']\nTask:Give milk to cat\nresponse:\n", "output": {"thoughts": "Grab milk and pour it into the plate so that the cat can drink", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[POINTAT] <cat>", "[PULL] <cat>", "[FIND] <milk>", "[GRAB] <milk>", "[FIND] <plate>", "[POUR] <milk> <plate>", "[TOUCH] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'blender', 'tablelamp', 'ground_coffee', 'bathroom', 'toothbrush', 'drawing', 'knife', 'orchid', 'desk', 'keyboard', 'brush', 'shower', 'homework', 'food_jam', 'pillow', 'dining_room', 'bowl', 'wall_clock', 'check', 'television', 'toothbrush_holder', 'tray', 'kettle', 'kitchen_counter', 'maindoor', 'sink', 'clothes_skirt', 'tooth_paste', 'dirt']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'drawing', 'knifeblock', 'floor', 'ceiling', 'ceilinglamp', 'mat', 'shower', 'picture', 'window', 'food_noodles', 'food_jam', 'wall', 'powersocket', 'oven', 'photoframe', 'check', 'television', 'toothbrush_holder', 'water', 'curtain', 'doorjamb', 'kitchen_counter', 'cd', 'headset', 'door', 'bathroom_cabinet', 'light', 'soap']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "go to a bathroom, open the shower, let water wash your body", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'detergent', 'drawing', 'knife', 'wallshelf', 'hanger', 'floor', 'ceiling', 'computer', 'mop_bucket', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'couch', 'clothes_hat', 'pillow', 'wall', 'box', 'photoframe', 'bed', 'coffe_maker', 'television', 'coin', 'doorjamb', 'door', 'sink']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room ,find sofa ,sit on sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_salt', 'microwave', 'chair', 'jelly', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'toaster', 'tvstand', 'faucet', 'shower', 'closetdrawer', 'candle', 'food_orange', 'wall', 'cup', 'oven_mitts', 'water', 'doorjamb', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'bag', 'creditcard']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Opening the kitchen cabinet to grab water glass and pour water from the faucet into it to drink.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[CLOSE] <kitchen_cabinet>", "[FIND] <sink>", "[PUTBACK] <water_glass> <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[GRAB] <water_glass>", "[FIND] <water>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'dog', 'toothbrush', 'bills', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'purse', 'faucet', 'balanceball', 'food_noodles', 'wall', 'bench', 'cup', 'photoframe', 'mouthwash', 'console', 'kitchen_counter', 'cupboard', 'sink', 'bag', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": go to bathroom. walk to sink,turn on faucet, put hands in water. rub hands together. grab soap. rub soap in hands. put soap down. put hands under water. rubs hands together. turn off faucet,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'detergent', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'character', 'stamp', 'keyboard', 'toaster', 'plate', 'faucet', 'tvstand', 'bowl', 'wall', 'oven', 'photoframe', 'coffe_maker', 'food_food', 'ice', 'curtain', 'shaving_cream', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the dishwasher and turn on the faucet. Rinse off the dishes. Put the dishes in the dishwasher. Put detergent in the dishwasher. Close the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <detergent>", "[GRAB] <detergent>", "[PUTBACK] <detergent> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'purse', 'mail', 'computer', 'laptop', 'deck_of_cards', 'faucet', 'foundation', 'mousepad', 'wall', 'wall_clock', 'bench', 'cup', 'doorjamb', 'mouse', 'door', 'light', 'table']\nTask:Sent email\nresponse:\n{\n    \"thoughts\": Walk into home office. Sit down in a chair at a desk. Turn on laptop. Bring up the internet. Go to gmail website. Begin composing an email. Hit the send button.,\n    \"actions\": ['[WALK] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <mail>', '[GRAB] <mail>', '[READ] <mail>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <laptop>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'cd_player', 'chair', 'walllamp', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'envelope', 'mail', 'button', 'toaster', 'ceilinglamp', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'nightstand', 'wall', 'video_game_controller', 'television', 'doorjamb', 'kitchen_counter', 'headset', 'hair', 'door', 'sink']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room, walk to cd player, turn on cd player, press play button, walk to sofa, sit in sofa and listen to music", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <button>", "[PUSH] <button>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'centerpiece', 'mouse', 'shoes', 'dresser', 'door', 'maindoor', 'sink', 'light', 'table', 'soap']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": go to the office, turn computer on, open an browser or any application to open any website.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'measuring_cup', 'microwave', 'toilet', 'dog', 'bathroom', 'knifeblock', 'bills', 'bathroom_counter', 'floor', 'table_cloth', 'toaster', 'plate', 'faucet', 'couch', 'towel', 'wall', 'bed', 'oven_mitts', 'water', 'shaving_cream', 'kitchen_counter', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'sink', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find sink. Walk to sink. Find soap. Put soap on sink. Pour water into hands. Scrub soap. Wash water.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[WALK] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <sink>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>", "[SCRUB] <soap>", "[WASH] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'wallshelf', 'cpuscreen', 'floor', 'shampoo', 'mop_bucket', 'toaster', 'mat', 'plate', 'shower', 'dish_soap', 'closetdrawer', 'dining_room', 'wall', 'photoframe', 'coffe_maker', 'colander', 'pencil', 'kitchen_counter', 'cupboard', 'clothes_dress', 'food_steak', 'home_office', 'dishwasher', 'soap', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'cat', 'drawing', 'wallshelf', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'bookshelf', 'computer', 'food_bacon', 'bookmark', 'brush', 'mat', 'mousepad', 'clothes_hat', 'wall', 'nightstand', 'powersocket', 'wall_clock', 'cup', 'oven_mitts', 'doorjamb', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "push any ligth button to turn it on, for example in the living room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'crayon', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'door', 'clothes_scarf', 'light']\nTask:Re arrange office\nresponse:\n{\n    \"thoughts\": Walk in.Turn on light.Move furniture. Change the room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[PUSH] <desk>', '[WALK] <chair>', '[PUSH] <chair>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'juice', 'food_ice_cream', 'microwave', 'knifeblock', 'orchid', 'floor', 'ceiling', 'character', 'bookmark', 'toaster', 'ceilinglamp', 'laptop', 'faucet', 'stereo', 'couch', 'wall', 'toy', 'cup', 'coin', 'ice', 'wine', 'curtain', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'clothes_dress', 'sink', 'instrument_guitar', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drink water from a cup", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[FIND] <faucet>", "[WALK] <sink>", "[PUTBACK] <cup> <sink>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[GRAB] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'freezer', 'trashcan', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'window', 'closetdrawer', 'clothes_pants', 'wall', 'bed', 'coffe_maker', 'television', 'tray', 'towel_rack', 'dresser', 'cupboard', 'hair', 'bathtub', 'sink', 'home_office', 'table', 'bathroom_cabinet', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Get some clothes together that make an appropriate outfit.  Put them on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'measuring_cup', 'jelly', 'chair', 'freezer', 'hanger', 'ceiling', 'thread', 'computer', 'mail', 'bookmark', 'deck_of_cards', 'faucet', 'mousepad', 'dining_room', 'broom', 'wall', 'wall_clock', 'box', 'razor', 'doorjamb', 'mouse', 'pencil', 'hairbrush', 'door', 'napkin', 'table', 'light', 'dirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter the Home office, then walk to the phone, pickup the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'water_glass', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'stovefan', 'mat', 'shower', 'closetdrawer', 'couch', 'wall', 'powersocket', 'oven', 'food_food', 'check', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen ,find fridge ,open fridge ,find water  glass , drink from water glass,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'fork', 'filing_cabinet', 'chair', 'cat', 'drawing', 'hanger', 'floor', 'character', 'tea_bag', 'mat', 'picture', 'electric_shaver', 'couch', 'food_jam', 'towel', 'dining_room', 'wall', 'oven', 'coffe_maker', 'tray', 'curtain', 'doorjamb', 'diary', 'cards', 'cupboard', 'door', 'bathroom_cabinet', 'light', 'tooth_paste']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up  the book open the book and read", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'child', 'photoframe', 'food_food', 'television', 'after_shave', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on tv.  Sit with kids.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[FIND] <child>', '[GREET] <child>', '[FIND] <child>', '[GREET] <child>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'orchid', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'computer', 'button', 'tvstand', 'stereo', 'window', 'closetdrawer', 'couch', 'candle', 'clothes_hat', 'pillow', 'nightstand', 'wall', 'cup', 'television', 'oven_mitts', 'curtain', 'centerpiece', 'coffee_pot', 'dresser', 'table', 'instrument_violin']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk into the Living room.  Find the television remote.  Use it to turn on the television.  Point the remote at the television. Press the channel button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'bathroom', 'detergent', 'bathroom_counter', 'button', 'toaster', 'brush', 'stereo', 'shower', 'faucet', 'food_orange', 'pillow', 'oven', 'photoframe', 'check', 'television', 'tray', 'kettle', 'mouse', 'cupboard', 'conditioner', 'door', 'home_office', 'table', 'dishwasher', 'clothes_skirt', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room, look for radio, grab on button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'filing_cabinet', 'chair', 'blow_dryer', 'ceiling', 'character', 'tea_bag', 'keyboard', 'piano_bench', 'ceilinglamp', 'mousepad', 'closetdrawer', 'dustpan', 'pillow', 'television', 'coffee_filter', 'drinking_glass', 'water', 'wine', 'curtain', 'pencil', 'vacuum_cleaner', 'controller', 'door', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'clothes_scarf']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. Take a glass out of the cabinet. Then I walk to the sink and turn the water on and fill the glass. Then I take a drink.", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <drinking_glass>", "[POUR] <water> <drinking_glass>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'trashcan', 'floor', 'tea_bag', 'thread', 'toaster', 'ceilinglamp', 'plate', 'dish_soap', 'wall', 'bed', 'coffe_maker', 'food_food', 'check', 'oven_mitts', 'tray', 'doorjamb', 'pencil', 'kitchen_counter', 'hairbrush', 'cupboard', 'hair', 'controller', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'tea', 'chair', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'plate', 'mousepad', 'couch', 'clothes_hat', 'wall', 'nightstand', 'cup', 'food_food', 'check', 'coffe_maker', 'after_shave', 'rag', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'hairdryer', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to living room,find the text book on the self ,sat on the chair ,open and read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'beer', 'toilet', 'chair', 'ceilingfan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'folder', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'microphone', 'wall', 'bench', 'light_bulb', 'television', 'curtain', 'food_steak', 'sink', 'napkin', 'home_office', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into my living room, find the remote and sit on my sofa. I look at the remote and switch on the television. I look at the Television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'mop', 'microwave', 'band-aids', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'phone', 'desk', 'clothes_jacket', 'toaster', 'faucet', 'picture', 'couch', 'clothes_hat', 'dining_room', 'wall', 'cup', 'coffe_maker', 'mouthwash', 'drinking_glass', 'kettle', 'kitchen_counter', 'paper_towel', 'cupboard', 'controller', 'laundry_detergent', 'sink', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "wipe loose dirt, place in the dishwasher, turn on the dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WIPE] <drinking_glass>", "[RINSE] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'soap']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to my home office, sit on the chair and turn on the computer. I open my email account to look for all the information I need.I pull up my online task on the computer to look at what the assigned task is.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'tea', 'knifeblock', 'cpuscreen', 'folder', 'bedroom', 'table_cloth', 'toaster', 'plate', 'homework', 'oil', 'closetdrawer', 'powersocket', 'tape', 'bed', 'novel', 'coffe_maker', 'television', 'razor', 'wine', 'doorjamb', 'shoes', 'dishrack', 'headset', 'cupboard', 'hairdryer', 'light', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk into living room, walk to stereo, turn on stereo, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'filing_cabinet', 'tablelamp', 'blender', 'mop', 'bathroom', 'detergent', 'toothbrush', 'bills', 'hanger', 'floor', 'bedroom', 'envelope', 'toaster', 'stovefan', 'clothes_socks', 'mousepad', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'toilet_paper', 'towel_rack', 'clothes_dress', 'bag', 'home_office', 'light', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'ground_coffee', 'walllamp', 'knife', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bedroom', 'button', 'stovefan', 'mat', 'shower', 'clothes_socks', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'bed', 'light_bulb', 'check', 'video_game_controller', 'mouthwash', 'food_onion', 'vacuum_cleaner', 'pajamas', 'cupboard', 'hair', 'instrument_guitar']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "go to bedroom, put pajamas on, lie in bed, sleep", "actions": ["[WALK] <bedroom>", "[WALK] <pajamas>", "[FIND] <pajamas>", "[GRAB] <pajamas>", "[PUTON] <pajamas>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'stovefan', 'balanceball', 'mat', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'light', 'dishwasher']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, sit at desk, turn on computer, enter password, open application and begin work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'drawing', 'floor', 'bedroom', 'bookshelf', 'stovefan', 'ceilinglamp', 'plate', 'dish_soap', 'towel', 'pillow', 'wall', 'cup', 'bed', 'food_food', 'check', 'coffe_maker', 'television', 'doorjamb', 'console', 'kitchen_counter', 'cupboard', 'door', 'clothes_scarf', 'table', 'dishwasher', 'clothes_skirt', 'soap', 'board_game']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go the in kitchen then grab the plate on the table. I put the plate and dish soap in the dishwasher. Finally, I turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'dough', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'tvstand', 'stereo', 'mousepad', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'sink', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to the bedroom and sit at my desk. I turn my radio on and listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'chair', 'dog', 'toothbrush', 'drawing', 'bills', 'ceiling', 'folder', 'desk', 'bedroom', 'remote_control', 'button', 'clothes_jacket', 'stovefan', 'mat', 'faucet', 'dining_room', 'broom', 'nightstand', 'wall', 'cutting_board', 'novel', 'mouse', 'kitchen_counter', 'cd', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pick up the remote. I press button on remote.", "actions": ["[WALK] <bedroom>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'chair', 'freezer', 'bathroom', 'trashcan', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'character', 'bedroom', 'computer', 'plate', 'stereo', 'window', 'food_kiwi', 'wall', 'nightstand', 'box', 'bed', 'clothes_gloves', 'coffe_maker', 'wooden_spoon', 'curtain', 'shoes', 'dishwasher', 'creditcard']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go into my bedroom and see my stereo. I decide to play music. I walk over to the stereo and turn it on. I sit down on the bed to listen to the music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[SWITCHON] <stereo>', '[FIND] <bed>', '[SIT] <bed>']\n}\n", "input": "\nScene_Objects:['glue', 'beer', 'measuring_cup', 'toilet', 'freezer', 'food_carrot', 'drawing', 'food_egg', 'hanger', 'floor', 'ceiling', 'bedroom', 'mat', 'laptop', 'shower', 'tvstand', 'closetdrawer', 'food_orange', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'drinking_glass', 'wine', 'doorjamb', 'door', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Robot Drinks the water", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>", "[PUTOBJBACK] <drinking_glass>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'shaving_cream', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": switch on the computer,  grab a chair and sit in front of the computer. type with keyboard and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['juice', 'toilet', 'freezer', 'band-aids', 'ground_coffee', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'mat', 'shower', 'couch', 'food_jam', 'wall', 'pasta', 'powersocket', 'oven', 'novel', 'coffe_maker', 'food_food', 'check', 'curtain', 'doorjamb', 'door', 'napkin', 'light', 'table', 'clothes_skirt']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Travel through the house and locate the bathroom, and proceed to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'freezer', 'bathroom', 'orchid', 'phone', 'dvd_player', 'character', 'bedroom', 'ceilinglamp', 'balanceball', 'iron', 'window', 'couch', 'food_orange', 'wall', 'tape', 'cup', 'basket_for_clothes', 'check', 'wooden_spoon', 'console', 'mouse', 'kitchen_counter', 'hairbrush', 'food_steak', 'sink', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'toilet', 'bathroom', 'knifeblock', 'drawing', 'cpuscreen', 'orchid', 'hanger', 'floor', 'bedroom', 'bookshelf', 'mop_bucket', 'stovefan', 'mat', 'homework', 'electric_shaver', 'clothes_socks', 'closetdrawer', 'food_orange', 'pillow', 'bowl', 'basket_for_clothes', 'cup', 'coffe_maker', 'tray', 'shaving_cream', 'cupboard', 'sink', 'board_game']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find Cat in living room.  Walk to cat.  Pet cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tea', 'chair', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'plate', 'mousepad', 'couch', 'clothes_hat', 'wall', 'nightstand', 'cup', 'food_food', 'check', 'coffe_maker', 'after_shave', 'rag', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'hairdryer', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": go to living room,find the text book on the self ,sat on the chair ,open and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'foundation', 'homework', 'closetdrawer', 'couch', 'pillow', 'broom', 'nightstand', 'wall', 'oven', 'light_bulb', 'television', 'curtain', 'dresser', 'bathroom_cabinet', 'table', 'dishwasher', 'creditcard']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'food_ice_cream', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'dvd_player', 'character', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'electric_shaver', 'mousepad', 'couch', 'form', 'wall', 'nightstand', 'clothes_gloves', 'television', 'curtain', 'headset', 'maindoor', 'table', 'clothes_shirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to Living room. Find Television. Walk to sofa. Sit on sofa. Find Remote. Grab remote. Turn to television. Look at television. Switch on Television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_pizza', 'chair', 'drawing', 'food_egg', 'hanger', 'floor', 'phone', 'ceiling', 'mail', 'table_cloth', 'stovefan', 'ceilinglamp', 'stereo', 'homework', 'couch', 'microphone', 'form', 'food_orange', 'bowl', 'oven', 'box', 'newspaper', 'tray', 'kitchen_counter', 'cupboard', 'controller', 'bathroom_cabinet', 'table', 'shoe_rack']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, walk to table, grab water glass and drink water", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>", "[PUTBACK] <water_glass> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'bathroom', 'milk', 'floor', 'bookshelf', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'bench', 'oven', 'toy', 'cutting_board', 'shelf', 'cup', 'standingmirror', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'console', 'pencil', 'comb', 'sink', 'home_office', 'scrabble']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Grab toys off floor.  Put it in a container.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'toilet', 'chair', 'freezer', 'walllamp', 'food_butter', 'drawing', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'toaster', 'nightstand', 'wall', 'powersocket', 'oven', 'pasta', 'box', 'food_cheese', 'photoframe', 'food_food', 'wooden_spoon', 'food_onion', 'tray', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'door', 'light', 'table', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and open the fridge. From the fridge take some chicken and close the fridge. Walk to the table and put the chicken on the table. Plug in oven and put the chicken into the oven.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <food_chicken> <table>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <table>", "[FIND] <oven>", "[PLUGIN] <oven>", "[OPEN] <oven>", "[GRAB] <food_chicken>", "[PUTIN] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'couch', 'wall', 'nightstand', 'toy', 'photoframe', 'curtain', 'rag', 'doorjamb', 'mouse', 'dresser', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": walk into Living room, look for computer, walk to computer, turn on computer, grab mouse, click mouse,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[WALK] <computer>', '[TURNTO] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>']\n}\n", "input": "\nScene_Objects:['microwave', 'board_game', 'chair', 'ground_coffee', 'detergent', 'knifeblock', 'floor', 'ceiling', 'alcohol', 'toaster', 'faucet', 'picture', 'closetdrawer', 'couch', 'dining_room', 'broom', 'wall', 'coffe_maker', 'check', 'water', 'drinking_glass', 'after_shave', 'rag', 'kitchen_counter', 'shoes', 'cupboard', 'hair', 'sink', 'table', 'tooth_paste']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Walk to faucet. Switch on faucet. Pour water into drinking glass. Switch off faucet. Drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <sink>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[SWITCHOFF] <faucet>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'folder', 'desk', 'alcohol', 'stovefan', 'plate', 'mousepad', 'wall', 'oven', 'food_food', 'check', 'tray', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'sink', 'clothes_scarf', 'light', 'clothes_shirt', 'teeth']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk over and pick up the phone,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'mousepad', 'closetdrawer', 'microphone', 'pillow', 'wall', 'bed', 'television', 'razor', 'doorjamb', 'kettle', 'mouse', 'address_book', 'door', 'light', 'table', 'dishwasher']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office. Walk to computer. Switch on computer. Open the address book.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <address_book>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'toaster', 'ceilinglamp', 'window', 'floor_lamp', 'couch', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'comb', 'door', 'bathtub', 'bathroom_cabinet', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Turning on lights,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'detergent', 'cat', 'food_butter', 'knife', 'phone', 'character', 'table_cloth', 'mat', 'deck_of_cards', 'stereo', 'form', 'pillow', 'dining_room', 'nightstand', 'wall', 'basket_for_clothes', 'photoframe', 'tray', 'doorjamb', 'kitchen_counter', 'chessboard', 'dresser', 'clothes_dress', 'table', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'fork', 'scissors', 'food_ice_cream', 'microwave', 'walllamp', 'knifeblock', 'knife', 'wallshelf', 'floor', 'ceiling', 'phone', 'alcohol', 'clothes_jacket', 'toaster', 'stovefan', 'faucet', 'couch', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'rag', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'colander']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Take out a rag. Turn on faucet and get water on the rag. Clean counter off.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <rag> <sink>', '[SWITCHOFF] <faucet>', '[GRAB] <rag>', '[SQUEEZE] <rag>', '[WALK] <kitchen_counter>', '[WIPE] <kitchen_counter>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['glue', 'toilet', 'freezer', 'wallshelf', 'hanger', 'bathroom_counter', 'bookshelf', 'mail', 'brush', 'mat', 'shoe-shine_kit', 'homework', 'closetdrawer', 'couch', 'food_orange', 'towel', 'dining_room', 'wall', 'powersocket', 'box', 'novel', 'check', 'coffe_maker', 'tray', 'doorjamb', 'diary', 'cupboard', 'home_office', 'bathroom_cabinet', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up the book, open to current page, read the words.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'dog', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'envelope', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'pasta', 'nightstand', 'bench', 'photoframe', 'tray', 'mouse', 'door', 'table', 'soap']\nTask:Work\nresponse:\n{\n    \"thoughts\": go to home office. turn the computer on. begin working,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'mop', 'bathroom', 'food_egg', 'clothes_underwear', 'bathroom_counter', 'phone', 'ceiling', 'computer', 'toaster', 'mat', 'shoe-shine_kit', 'shower', 'clothes_hat', 'clothes_pants', 'food_orange', 'food_kiwi', 'basket_for_clothes', 'coffe_maker', 'television', 'curtain', 'mouse', 'cd', 'comb', 'clothes_dress', 'clothes_skirt', 'soap', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I need to get the dirty clothes basket.  Walk to bathroom.  Put all the clothes in washing machine.  Add soap.  Turn on machine.", "actions": ["[WALK] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_cake', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'couch', 'wall', 'powersocket', 'bed', 'television', 'doorjamb', 'mouse', 'hairdryer', 'door', 'sink', 'home_office', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Using desktop computer in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'food_bread', 'scissors', 'microwave', 'jelly', 'detergent', 'toothbrush', 'drawing', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'envelope', 'shower', 'faucet', 'window', 'candle', 'pillow', 'wall', 'tape', 'keys', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'sink', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone and pick it up.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'scissors', 'measuring_cup', 'toilet', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'remote_control', 'purse', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'form', 'dining_room', 'wall', 'powersocket', 'television', 'curtain', 'food_onion', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to living room. Find remote control. Look at remote control. Push power on button. Find sofa. Sit down on sofa. Watch Television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'cat', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'clothes_jacket', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'couch', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'box', 'food_food', 'television', 'coffee_pot', 'kettle', 'cards', 'mouse', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'crayon', 'chair', 'cat', 'dough', 'milk', 'hanger', 'floor', 'keyboard', 'mail', 'table_cloth', 'stovefan', 'plate', 'faucet', 'tvstand', 'window', 'couch', 'form', 'dining_room', 'cutting_board', 'photoframe', 'check', 'wine', 'dresser', 'hairbrush', 'pajamas', 'conditioner', 'home_office', 'bathroom_cabinet', 'table']\nTask:Give milk to cat\nresponse:\n{\n    \"thoughts\": Grab milk and pour it into the plate so that the cat can drink,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[POINTAT] <cat>', '[PULL] <cat>', '[FIND] <milk>', '[GRAB] <milk>', '[FIND] <plate>', '[POUR] <milk> <plate>', '[TOUCH] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'ground_coffee', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'toaster', 'plate', 'faucet', 'foundation', 'mousepad', 'food_jam', 'food_kiwi', 'wall', 'pasta', 'powersocket', 'food_food', 'coffe_maker', 'rag', 'shaving_cream', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Take a dish and put it in the dishwasher to be washed", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'computer', 'tvstand', 'faucet', 'stereo', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'wall', 'powersocket', 'shelf', 'food_food', 'oven_mitts', 'curtain', 'doorjamb', 'coffee_pot', 'mouse', 'vacuum_cleaner', 'dresser', 'light', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": approach the light switch, raise hand to the switch, flip the switch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'floor', 'ceiling', 'computer', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'basket_for_clothes', 'food_food', 'check', 'television', 'tray', 'curtain', 'dresser', 'door', 'home_office', 'table', 'clothes_shirt', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I reach for the remote. I push the on button on the remote. I sit down on the couch and watch television.", "actions": ["[WALK] <home_office>", "[WALK] <button>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'wallshelf', 'bookshelf', 'clothes_jacket', 'toaster', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'dining_room', 'nightstand', 'wall', 'wall_clock', 'powersocket', 'basket_for_clothes', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'console', 'dishrack', 'hairdryer', 'door', 'bathtub']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'knife', 'desk', 'bookshelf', 'button', 'clothes_jacket', 'stovefan', 'deck_of_cards', 'mousepad', 'closetdrawer', 'oil', 'towel', 'pillow', 'broom', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'television', 'wine', 'curtain', 'after_shave', 'cards', 'home_office', 'light', 'clothes_skirt', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[WALK] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'blow_dryer', 'walllamp', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'faucet', 'picture', 'iron', 'mousepad', 'bowl', 'wall', 'powersocket', 'bench', 'photoframe', 'toilet_paper', 'coffe_maker', 'toothbrush_holder', 'tray', 'doorjamb', 'dishrack', 'sink', 'bathroom_cabinet', 'home_office', 'light']\nTask:Get toilet paper\nresponse:\n{\n    \"thoughts\": There are extra toilet papers in the bottom cabinet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <toilet_paper>', '[GRAB] <toilet_paper>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['scissors', 'freezer', 'walllamp', 'knifeblock', 'wallshelf', 'clothes_underwear', 'ceiling', 'bookshelf', 'keyboard', 'ceilinglamp', 'plate', 'stereo', 'clothes_socks', 'clothes_pants', 'pillow', 'bowl', 'powersocket', 'cutting_board', 'basket_for_clothes', 'razor', 'coin', 'wine', 'curtain', 'pencil', 'kitchen_counter', 'cd', 'dresser', 'laundry_detergent', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'bathroom', 'detergent', 'cat', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'mousepad', 'form', 'wall', 'powersocket', 'photoframe', 'food_food', 'check', 'coffee_filter', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'controller', 'door', 'light', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['fork', 'tea', 'jelly', 'dog', 'chair', 'ground_coffee', 'bathroom', 'phone', 'bedroom', 'stovefan', 'mat', 'plate', 'faucet', 'shoe-shine_kit', 'wall', 'nightstand', 'powersocket', 'cup', 'basket_for_clothes', 'newspaper', 'food_food', 'coffe_maker', 'tray', 'curtain', 'food_steak', 'door', 'bag', 'instrument_guitar', 'light', 'dishwasher']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <home_office>", "[DROP] <plate>", "[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <home_office>", "[DROP] <cup>", "[WALK] <dining_room>", "[WALK] <fork>", "[FIND] <fork>", "[GRAB] <fork>", "[WALK] <home_office>", "[DROP] <fork>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'jelly', 'freezer', 'walllamp', 'bathroom', 'hanger', 'folder', 'bedroom', 'desk', 'tvstand', 'window', 'food_jam', 'microphone', 'dining_room', 'wall_clock', 'powersocket', 'bench', 'shelf', 'bed', 'video_game_controller', 'food_donut', 'water', 'doorjamb', 'cards', 'towel_rack', 'mouse', 'sink', 'colander', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the water, wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <water>', '[FIND] <water>', '[TURNTO] <water>', '[LOOKAT] <water>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['dog', 'freezer', 'bathroom', 'food_carrot', 'knifeblock', 'food_egg', 'hanger', 'floor', 'phone', 'stovefan', 'mat', 'plate', 'stereo', 'dish_soap', 'food_noodles', 'closetdrawer', 'mousepad', 'dustpan', 'cup', 'food_food', 'coffe_maker', 'check', 'oven_mitts', 'water', 'food_rice', 'curtain', 'mouse', 'hairbrush', 'sink', 'home_office']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen, fins soap, arrange dishes, wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[FIND] <dish_soap>", "[FIND] <water>", "[GRAB] <plate>", "[GRAB] <dish_soap>", "[WASH] <plate>", "[RINSE] <plate>", "[DROP] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'curtain', 'shoes', 'cd', 'vacuum_cleaner', 'dresser', 'cupboard', 'clothes_dress', 'home_office', 'table', 'shoe_rack']\nTask:Take off shoes\nresponse:\n{\n    \"thoughts\": Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'scissors', 'food_ice_cream', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'stamp', 'piano_bench', 'ceilinglamp', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'toothbrush_holder', 'curtain', 'diary', 'vacuum_cleaner', 'sink', 'home_office', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'jelly', 'bills', 'food_egg', 'hanger', 'floor', 'phone', 'ceiling', 'dvd_player', 'mat', 'window', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'wall', 'cutting_board', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'hair', 'table', 'soap', 'clothes_shirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to living room.  Find sofa.  Sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'dog', 'detergent', 'needle', 'hanger', 'floor', 'ceiling', 'envelope', 'table_cloth', 'ceilinglamp', 'broom', 'wall', 'cutting_board', 'photoframe', 'bed', 'coffe_maker', 'check', 'coffee_filter', 'water', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'controller', 'table', 'dishwasher', 'clothes_skirt', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I find the coffee machine, fill it with water and coffee then press the button to make the coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'bed', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": Type on a computer in the home office.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'food_carrot', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'remote_control', 'computer', 'button', 'stovefan', 'tvstand', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'newspaper', 'toilet_paper', 'television', 'tray', 'curtain', 'mouse', 'dresser', 'sink', 'bathroom_cabinet', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'nail_polish', 'bathroom', 'cat', 'dough', 'toothbrush', 'desk', 'button', 'stovefan', 'shower', 'faucet', 'window', 'pillow', 'nightstand', 'wall_clock', 'oven', 'cutting_board', 'novel', 'check', 'television', 'toothbrush_holder', 'lighter', 'kettle', 'chessboard', 'dresser', 'sink', 'home_office', 'clothes_skirt']\nTask:Keep cats out of room\nresponse:\n{\n    \"thoughts\": Stand in the doorway and create a loud noise so the cats will stay out.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'toilet', 'freezer', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'character', 'desk', 'food_cake', 'toaster', 'ceilinglamp', 'couch', 'pillow', 'powersocket', 'basket_for_clothes', 'cup', 'newspaper', 'food_food', 'check', 'oven_mitts', 'wooden_spoon', 'centerpiece', 'kitchen_counter', 'door', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office, find where the phone is and lift the reciever.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'bathroom', 'trashcan', 'ceilingfan', 'bills', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'folder', 'stovefan', 'tvstand', 'window', 'homework', 'wall', 'oven', 'powersocket', 'cutting_board', 'basket_for_clothes', 'food_cheese', 'clothes_gloves', 'check', 'television', 'toothbrush_holder', 'curtain', 'bathroom_cabinet', 'laser_pointer']\nTask:Open window\nresponse:\n{\n    \"thoughts\": Unlock the latch at the top. Push the bottom half of the window upwards.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <window>', '[OPEN] <window>']\n}\n", "input": "\nScene_Objects:['sheets', 'fork', 'pot', 'tablelamp', 'measuring_cup', 'freezer', 'detergent', 'cat', 'bathroom_counter', 'keyboard', 'purse', 'bookshelf', 'toaster', 'brush', 'faucet', 'stereo', 'homework', 'toy', 'box', 'basket_for_clothes', 'coin', 'curtain', 'rag', 'centerpiece', 'dishrack', 'clothes_dress', 'bag', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the washing machine. Open the washing machine. Find the basket for clothes. Find the clothes_dress. Grab the clothes_dress. Put the clothes_dress into the washing machine. Find the clothes_shirt. Grab the clothes_shirt. Put the clothes_shirt into the washing machine. Close the washing machine. Switch on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <basket_for_clothes>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'drawing', 'knifeblock', 'floor', 'ceiling', 'ceilinglamp', 'mat', 'shower', 'picture', 'window', 'food_noodles', 'food_jam', 'wall', 'powersocket', 'oven', 'photoframe', 'check', 'television', 'toothbrush_holder', 'water', 'curtain', 'doorjamb', 'kitchen_counter', 'cd', 'headset', 'door', 'bathroom_cabinet', 'light', 'soap']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'computer', 'mat', 'shoe-shine_kit', 'oil', 'pillow', 'wall', 'pasta', 'powersocket', 'cup', 'food_food', 'video_game_controller', 'toothbrush_holder', 'television', 'doorjamb', 'kettle', 'door', 'light', 'clothes_skirt', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on light in an area", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'ground_coffee', 'knifeblock', 'hanger', 'floor', 'ceiling', 'purse', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'wall', 'cup', 'razor', 'coin', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": To go kitchen and find a cup. open the faucet and fill water in the cup and drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <cup>', '[GRAB] <cup>', '[TURNTO] <faucet>', '[POINTAT] <faucet>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'tea', 'measuring_cup', 'walllamp', 'orchid', 'floor', 'ceiling', 'bookshelf', 'computer', 'tvstand', 'window', 'electric_shaver', 'oil', 'closetdrawer', 'couch', 'mousepad', 'pillow', 'nightstand', 'wall', 'tape', 'bed', 'television', 'curtain', 'rag', 'mouse', 'shoes', 'dresser', 'cupboard', 'controller', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch TV in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'scissors', 'food_ice_cream', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'stamp', 'piano_bench', 'ceilinglamp', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'toothbrush_holder', 'curtain', 'diary', 'vacuum_cleaner', 'sink', 'home_office', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'microwave', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'food_egg', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'toaster', 'shower', 'faucet', 'oil', 'food_kiwi', 'nightstand', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'food_food', 'oven_mitts', 'tray', 'doorjamb', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Open fridge. Take food from fridge. Close fridge. Put food on kitchen counter. Prepare food. Turn on stove. Open kitchen cabinet. Take pot. Close cabinet. Put food in pot. Put pot on stove. Cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <kitchen_counter>", "[PUTBACK] <food_food> <kitchen_counter>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'character', 'video_game_console', 'bedroom', 'computer', 'button', 'table_cloth', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'box', 'video_game_controller', 'television', 'curtain', 'pencil', 'dresser', 'door', 'table']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tea', 'chair', 'detergent', 'toothbrush', 'orchid', 'knifeblock', 'cpuscreen', 'kitchen_cabinet', 'floor', 'phone', 'bedroom', 'keyboard', 'ceilinglamp', 'faucet', 'window', 'dish_soap', 'food_jam', 'food_orange', 'pillow', 'wall', 'oven', 'bed', 'novel', 'coffe_maker', 'kitchen_counter', 'dishrack', 'cupboard', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTIN] <dish_soap> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'floor', 'ceiling', 'computer', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'basket_for_clothes', 'food_food', 'check', 'television', 'tray', 'curtain', 'dresser', 'door', 'home_office', 'table', 'clothes_shirt', 'coffee']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room. I reach for the remote. I push the on button on the remote. I sit down on the couch and watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <button>', '[FIND] <button>', '[PUSH] <button>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'jelly', 'bathroom', 'knife', 'dvd_player', 'character', 'bedroom', 'thread', 'computer', 'toaster', 'shoe-shine_kit', 'deck_of_cards', 'shower', 'electric_shaver', 'homework', 'food_peanut_butter', 'form', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'novel', 'light_bulb', 'food_food', 'curtain', 'coffee_pot', 'food_steak', 'door', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the radio. Turn on the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'orchid', 'ceiling', 'desk', 'computer', 'food_bacon', 'stovefan', 'ceilinglamp', 'plate', 'shower', 'faucet', 'window', 'form', 'dining_room', 'oven', 'basket_for_clothes', 'keys', 'novel', 'food_food', 'coffe_maker', 'food_rice', 'tray', 'doorjamb', 'cards', 'mouse', 'pajamas', 'home_office', 'light', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the playback device you want to listen to. turn on. press play. enjoyu music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['scissors', 'mop', 'bathroom', 'food_egg', 'wallshelf', 'clothes_underwear', 'character', 'bookshelf', 'food_cake', 'button', 'bookmark', 'laptop', 'mat', 'iron', 'wall', 'powersocket', 'television', 'console', 'mouse', 'shoes', 'pajamas', 'cupboard', 'door', 'hairdryer', 'home_office', 'light', 'bathroom_cabinet', 'soap', 'laser_pointer', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.", "actions": ["[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'ground_coffee', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'powersocket', 'check', 'television', 'doorjamb', 'mouse', 'pencil', 'dresser', 'sink', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to the home office to work on my project. I turn on the light and walk to my desk. I turn on the computer. Then I look through the documents. After I'm done, I turn off the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <check>', '[TURNTO] <check>', '[LOOKAT] <check>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_oatmeal', 'toilet', 'freezer', 'food_carrot', 'drawing', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'stamp', 'food_cake', 'computer', 'mat', 'shower', 'food_cereal', 'wall', 'nightstand', 'powersocket', 'food_vegetable', 'photoframe', 'newspaper', 'food_food', 'coffe_maker', 'doorjamb', 'pencil', 'door', 'bathroom_cabinet', 'light', 'scrabble']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_vegetable> <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTIN] <food_carrot> <freezer>", "[FIND] <food_cereal>", "[GRAB] <food_cereal>", "[PUTIN] <food_cereal> <freezer>", "[FIND] <food_oatmeal>", "[GRAB] <food_oatmeal>", "[PUTIN] <food_oatmeal> <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[FIND] <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'ground_coffee', 'blow_dryer', 'orchid', 'wallshelf', 'bathroom_counter', 'character', 'desk', 'bookshelf', 'alcohol', 'computer', 'ceilinglamp', 'laptop', 'faucet', 'pillow', 'wall_clock', 'powersocket', 'oven', 'cutting_board', 'cup', 'television', 'oven_mitts', 'after_shave', 'doorjamb', 'diary', 'towel_rack', 'cards', 'maindoor']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find cat. Walk to cat. Pull cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[PULL] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'bathroom', 'bills', 'orchid', 'cpuscreen', 'floor', 'phone', 'character', 'bedroom', 'piano_bench', 'toaster', 'mat', 'ceilinglamp', 'window', 'foundation', 'electric_shaver', 'oil', 'closetdrawer', 'couch', 'broom', 'nightstand', 'wall', 'cup', 'bed', 'check', 'toothbrush_holder', 'sink', 'napkin', 'dishwasher']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <sheets>", "[PULL] <sheets>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'microwave', 'toilet', 'freezer', 'ground_coffee', 'longboard', 'shampoo', 'character', 'thread', 'food_bacon', 'piano_bench', 'balanceball', 'stereo', 'window', 'picture', 'mousepad', 'pillow', 'toy', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'oven_mitts', 'colander', 'doorjamb', 'shaving_cream', 'dresser', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[WALK] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['water_glass', 'tablelamp', 'toilet', 'chair', 'walllamp', 'detergent', 'cat', 'knife', 'cpuscreen', 'floor', 'ceiling', 'mail', 'toaster', 'ceilinglamp', 'stovefan', 'stereo', 'homework', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'tape', 'newspaper', 'photoframe', 'food_food', 'television', 'water', 'tray', 'bathroom_cabinet']\nTask:Drink\nresponse:\n", "output": {"thoughts": "GOING TO DRINK WATER IN KITCHEN", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'chair', 'freezer', 'cpuscreen', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'stovefan', 'mousepad', 'wall', 'keys', 'check', 'oven_mitts', 'doorjamb', 'mouse', 'pencil', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'napkin', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": turn on the light for room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[SWITCHON] <light>', '[FIND] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'toilet', 'freezer', 'bathroom', 'drawing', 'floor', 'ceiling', 'character', 'computer', 'table_cloth', 'mat', 'shower', 'window', 'food_orange', 'broom', 'nightstand', 'wall', 'powersocket', 'basket_for_clothes', 'photoframe', 'bed', 'food_food', 'doorjamb', 'hairbrush', 'controller', 'door', 'sink', 'napkin', 'light']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Go kitchen and make something for eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'bathroom', 'toothbrush', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'phone', 'bedroom', 'computer', 'button', 'toaster', 'mat', 'stereo', 'foundation', 'window', 'couch', 'bench', 'oven', 'cup', 'check', 'television', 'tray', 'dresser', 'home_office', 'light', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room, look for radio, grab on button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'chair', 'milk', 'toothbrush', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'alcohol', 'mop_bucket', 'table_cloth', 'mat', 'faucet', 'homework', 'clothes_socks', 'closetdrawer', 'teeth', 'mousepad', 'wall', 'powersocket', 'doorjamb', 'shaving_cream', 'sink', 'bathroom_cabinet', 'light', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_pants', 'form', 'pillow', 'towel', 'wall', 'powersocket', 'bench', 'newspaper', 'television', 'curtain', 'doorjamb', 'mouse', 'hair', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'food_bread', 'filing_cabinet', 'scissors', 'knife', 'character', 'stamp', 'bookshelf', 'keyboard', 'bookmark', 'ceilinglamp', 'tvstand', 'shower', 'foundation', 'food_noodles', 'towel', 'tape', 'check', 'mouthwash', 'television', 'wooden_spoon', 'tray', 'kettle', 'kitchen_counter', 'cd', 'laundry_detergent', 'sink', 'light', 'soap', 'instrument_violin']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and get the soap and wash my both hand using the soap and rinse my hands and use towel.", "actions": ["[WALK] <dining_room>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[FIND] <towel>", "[RINSE] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'tea', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'dresser', 'door', 'home_office']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Keep my keyboard vacuumed and free of debris.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'food_ice_cream', 'chair', 'walllamp', 'knifeblock', 'hanger', 'floor', 'ceiling', 'food_bacon', 'toaster', 'ceilinglamp', 'oil', 'closetdrawer', 'couch', 'food_jam', 'pillow', 'wall', 'nightstand', 'newspaper', 'television', 'coin', 'lighter', 'laundry_detergent', 'napkin', 'home_office', 'dishwasher', 'soap', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, find sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'knifeblock', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'television', 'curtain', 'doorjamb', 'cd', 'pajamas', 'door', 'table', 'light', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sitting on the couch watching my favorite television show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['water_glass', 'blender', 'food_salt', 'chair', 'walllamp', 'nail_polish', 'bathroom', 'milk', 'drawing', 'wallshelf', 'clothes_underwear', 'bathroom_counter', 'floor', 'character', 'stamp', 'envelope', 'button', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'closetdrawer', 'form', 'coffe_maker', 'water', 'mouse', 'dresser', 'hairbrush', 'cupboard', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, hold drinking glass, lift glass, pour water in drinking glass", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[FIND] <water>", "[GRAB] <water>", "[WALK] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'bed', 'coin', 'curtain', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'light']\nTask:Wipe down baseboards please\nresponse:\n{\n    \"thoughts\": Use a rag to get all the dust off the baseboards please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <wall>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'walllamp', 'detergent', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'stovefan', 'toaster', 'faucet', 'dining_room', 'wall', 'pasta', 'oven', 'cup', 'food_cheese', 'keys', 'coffe_maker', 'water', 'tray', 'centerpiece', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'clothes_shirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, walk to cupboard, open cupboard, grab cup, walk to sink, pour water in cup, drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'walllamp', 'orchid', 'needle', 'clothes_underwear', 'dvd_player', 'character', 'desk', 'mop_bucket', 'table_cloth', 'brush', 'ceilinglamp', 'mat', 'stereo', 'dining_room', 'nightstand', 'powersocket', 'toy', 'shelf', 'check', 'toothbrush_holder', 'mouse', 'chessboard', 'dresser', 'laundry_detergent', 'home_office', 'bathroom_cabinet', 'teeth', 'creditcard']\nTask:Put toys away\nresponse:\n{\n    \"thoughts\": Pick up all the toys on the floor and put them in their correct storage bin or shelf.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'crayon', 'toilet', 'chair', 'freezer', 'milk', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'brush', 'mat', 'shower', 'nightstand', 'wall', 'powersocket', 'tape', 'food_cheese', 'photoframe', 'food_food', 'check', 'doorjamb', 'cupboard', 'controller', 'door', 'clothes_scarf', 'light', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I pick up groceries. I open fridge. I put groceries in the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[DROP] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'walllamp', 'drawing', 'orchid', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'couch', 'candle', 'pillow', 'wall', 'nightstand', 'powersocket', 'keys', 'television', 'curtain', 'doorjamb', 'door', 'sink', 'table']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Sit down on couch in front of TV.  Use remote to turn on the TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'toilet', 'band-aids', 'food_carrot', 'dough', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'folder', 'bedroom', 'shoe-shine_kit', 'closetdrawer', 'food_orange', 'nightstand', 'wall', 'powersocket', 'coin', 'shaving_cream', 'pencil', 'cupboard', 'sink', 'napkin', 'instrument_guitar', 'bathroom_cabinet', 'soap', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'water_glass', 'pot', 'freezer', 'walllamp', 'bathroom', 'food_butter', 'food_carrot', 'milk', 'knifeblock', 'knife', 'floor', 'ceiling', 'mail', 'computer', 'stovefan', 'ceilinglamp', 'brush', 'faucet', 'wall', 'nightstand', 'cutting_board', 'cup', 'photoframe', 'water', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'soap']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I enter the kitchen and grab a waterglass from the cupboard. I open the fridge and grap some water. I pour the water into the cup and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[OPEN] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'chair', 'knifeblock', 'hanger', 'floor', 'toaster', 'stovefan', 'shoe-shine_kit', 'faucet', 'homework', 'food_noodles', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'cup', 'newspaper', 'coffee_filter', 'food_rice', 'console', 'mouse', 'kitchen_counter', 'dishrack', 'comb', 'cupboard', 'door', 'sink']\nTask:Drink\nresponse:\n", "output": {"thoughts": "go to the kitchen, walk to the cabinet, pic up a glass. carry the glass to the sink. turn on the water and fill up the glass. drink the water.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'table_cloth', 'bookmark', 'mat', 'shower', 'faucet', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'oven', 'keys', 'coffe_maker', 'television', 'tray', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'hairdryer', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I open the room door and press the light button on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to the home office to find the computer. I turn switch computer on , i open the email. i write the email and  send it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'ceilingfan', 'dough', 'milk', 'knifeblock', 'bills', 'floor', 'ceiling', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'mop_bucket', 'ceilinglamp', 'microphone', 'wall', 'bench', 'powersocket', 'clothes_gloves', 'light_bulb', 'food_food', 'coffe_maker', 'tray', 'curtain', 'shaving_cream', 'controller', 'door', 'maindoor', 'dirt']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Go to kitchen and put groceries in fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'floor', 'ceiling', 'keyboard', 'food_cake', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'stereo', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'cup', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'food_steak', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I sit on the sofa. I put my legs on the sofa and recline.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'purse', 'toaster', 'faucet', 'shower', 'homework', 'pillow', 'wall', 'oven', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to kitchen, turn on water, find soap, apply to hand, lather, rinse, dry hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'ground_coffee', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'keyboard', 'mat', 'tvstand', 'plate', 'pillow', 'wall', 'nightstand', 'bed', 'coffe_maker', 'television', 'coffee_filter', 'razor', 'tray', 'kitchen_counter', 'cupboard', 'hairdryer', 'home_office', 'light', 'dishwasher', 'table', 'dirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Make coffee in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'milk', 'floor', 'ceiling', 'desk', 'keyboard', 'button', 'mop_bucket', 'toaster', 'stereo', 'window', 'wall', 'powersocket', 'bench', 'newspaper', 'coffe_maker', 'tray', 'curtain', 'food_sugar', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'coffee_cup', 'dishwasher', 'light', 'clothes_shirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[LOOKAT] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <milk>', '[GRAB] <milk>', '[PUTBACK] <milk> <coffe_maker>', '[FIND] <food_sugar>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[SWITCHON] <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'blender', 'chair', 'bathroom', 'hanger', 'floor', 'tea_bag', 'food_cake', 'food_bacon', 'piano_bench', 'mat', 'plate', 'dish_soap', 'mousepad', 'closetdrawer', 'couch', 'clothes_socks', 'wall', 'powersocket', 'coffe_maker', 'curtain', 'console', 'kitchen_counter', 'cupboard', 'door', 'sink', 'napkin', 'dishwasher', 'colander']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'band-aids', 'toothbrush', 'orchid', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'faucet', 'picture', 'shower', 'mousepad', 'candle', 'food_orange', 'oven', 'basket_for_clothes', 'bed', 'check', 'food_food', 'television', 'tray', 'after_shave', 'headset', 'comb', 'dresser', 'laundry_detergent', 'napkin', 'bathroom_cabinet']\nTask:Sit in chair\nresponse:\n{\n    \"thoughts\": I sit in a chair to rest and eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'food_carrot', 'orchid', 'floor', 'ceiling', 'shampoo', 'bedroom', 'remote_control', 'computer', 'button', 'ceilinglamp', 'tvstand', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'pasta', 'oven', 'novel', 'television', 'curtain', 'pencil', 'dresser', 'sink', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'detergent', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'purse', 'toaster', 'mat', 'faucet', 'iron', 'mousepad', 'wall', 'powersocket', 'cup', 'coffe_maker', 'water', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'clothes_scarf', 'colander']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'walllamp', 'bathroom', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'shoe-shine_kit', 'faucet', 'window', 'mousepad', 'couch', 'wall', 'oven', 'newspaper', 'coffe_maker', 'food_food', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'door', 'sink', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'chair', 'walllamp', 'toothbrush', 'wallshelf', 'dvd_player', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'mousepad', 'couch', 'form', 'pillow', 'bowl', 'nightstand', 'wall_clock', 'bench', 'oven', 'powersocket', 'clothes_gloves', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'hairdryer', 'bathtub', 'home_office', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I go to my bed room and sit on the chair and in front of me is my desktop computer and start writing an email.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'band-aids', 'freezer', 'food_carrot', 'toothbrush', 'orchid', 'floor', 'mail', 'button', 'faucet', 'foundation', 'oil', 'pillow', 'towel', 'food_food', 'toothbrush_holder', 'doorjamb', 'lighter', 'kettle', 'headset', 'cd', 'dresser', 'hairbrush', 'controller', 'napkin', 'light', 'dishwasher', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, take toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now brush teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[TURNTO] <toothbrush>", "[LOOKAT] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'dog', 'cpuscreen', 'keyboard', 'ceilinglamp', 'plate', 'laptop', 'faucet', 'window', 'stereo', 'oil', 'food_noodles', 'mousepad', 'dining_room', 'nightstand', 'food_food', 'check', 'wine', 'curtain', 'after_shave', 'comb', 'dresser', 'napkin', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I reach out arm. I put hand on radio. I turn on radio. I pick a station.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[TOUCH] <stereo>', '[TURNTO] <stereo>']\n}\n", "input": "\nScene_Objects:['sheets', 'detergent', 'orchid', 'wallshelf', 'floor', 'ceiling', 'tea_bag', 'stovefan', 'toaster', 'faucet', 'shower', 'foundation', 'tooth_paste', 'closetdrawer', 'candle', 'nightstand', 'cup', 'coffe_maker', 'food_food', 'television', 'coffee_filter', 'tray', 'curtain', 'rag', 'clothes_scarf', 'table', 'soap', 'instrument_violin', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'microwave', 'chair', 'freezer', 'detergent', 'food_egg', 'hanger', 'floor', 'ceiling', 'toaster', 'plate', 'shower', 'mousepad', 'broom', 'wall', 'pasta', 'keys', 'novel', 'coffe_maker', 'toothbrush_holder', 'coffee_filter', 'wooden_spoon', 'towel_rack', 'controller', 'maindoor', 'sink', 'dishwasher', 'coffee', 'creditcard']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'toothbrush', 'drawing', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'mat', 'faucet', 'shower', 'food_noodles', 'form', 'wall', 'wall_clock', 'oven', 'tape', 'photoframe', 'keys', 'toothbrush_holder', 'rag', 'doorjamb', 'sink', 'bathroom_cabinet', 'instrument_guitar', 'dishwasher', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I walk in to the bathroom. I turn on the faucet. I find the toothbrush and toothpaste. I grab the toothbrush and toothpaste. I rinse off the toothbrush. I add toothpaste to the toothbrush. I put back the toothpaste. I wash my teeth with the toothbrush. I rinse off the toothbrush. I put the toothbrush back. I rinse my mouth out. I turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[RINSE] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>", "[PUTOBJBACK] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>", "[RINSE] <toothbrush>", "[PUTOBJBACK] <toothbrush>", "[RINSE] <teeth>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'beer', 'measuring_cup', 'band-aids', 'woman', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'phone', 'ceiling', 'character', 'desk', 'man', 'thread', 'clothes_jacket', 'tvstand', 'stereo', 'faucet', 'window', 'shower', 'couch', 'form', 'cutting_board', 'novel', 'television', 'bathroom_cabinet', 'table', 'dishwasher', 'instrument_violin']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'mat', 'deck_of_cards', 'shower', 'tvstand', 'food_jam', 'dining_room', 'wall', 'nightstand', 'powersocket', 'food_cheese', 'photoframe', 'novel', 'food_food', 'toilet_paper', 'coffe_maker', 'food_donut', 'doorjamb', 'kitchen_counter', 'dresser', 'door', 'light', 'tooth_paste', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "in my home i walk to the kitchen with some groceries and locate the fridge,then open the fridge put the groceries into the fridge and closed.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'bed', 'coin', 'curtain', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'light']\nTask:Wipe down baseboards please\nresponse:\n{\n    \"thoughts\": Use a rag to get all the dust off the baseboards please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <wall>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'food_salt', 'jelly', 'walllamp', 'detergent', 'knife', 'drawing', 'hanger', 'floor', 'dvd_player', 'character', 'desk', 'bedroom', 'plate', 'dish_soap', 'oil', 'wall', 'powersocket', 'bed', 'light_bulb', 'coffe_maker', 'toilet_paper', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Put dishes in the dishwasher and start the dishwasher.", "actions": ["[RUN] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[WALK] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'beer', 'microwave', 'chair', 'bathroom', 'trashcan', 'dough', 'knifeblock', 'floor', 'computer', 'stovefan', 'mat', 'balanceball', 'faucet', 'window', 'couch', 'candle', 'wall', 'toy', 'bed', 'photoframe', 'coffe_maker', 'video_game_controller', 'wooden_spoon', 'curtain', 'dresser', 'pajamas', 'hair', 'maindoor']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go to bedroom, put pajamas on, lie in bed, sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <pajamas>', '[FIND] <pajamas>', '[GRAB] <pajamas>', '[PUTON] <pajamas>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'bathroom', 'orchid', 'floor', 'ceiling', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'food_peanut_butter', 'pillow', 'bowl', 'dining_room', 'nightstand', 'wall', 'powersocket', 'cup', 'check', 'television', 'curtain', 'dresser', 'cupboard', 'home_office', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'blender', 'microwave', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'keyboard', 'toaster', 'laptop', 'stereo', 'window', 'food_peanut_butter', 'dining_room', 'broom', 'nightstand', 'oven', 'cup', 'photoframe', 'bed', 'television', 'after_shave', 'curtain', 'pencil', 'bathroom_cabinet', 'light', 'dishwasher', 'vase']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Clean the welcome mat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <vacuum_cleaner>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'chair', 'dough', 'wallshelf', 'hanger', 'phone', 'tea_bag', 'bookshelf', 'thread', 'stovefan', 'mat', 'tvstand', 'faucet', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'oven', 'cup', 'food_donut', 'centerpiece', 'mouse', 'vacuum_cleaner', 'chessboard', 'laundry_detergent', 'sink', 'home_office', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to bedroom and grab dirty pants and take to the washing machine in entrance hall. Put clothes in washing machine and then put in laundry detergent.", "actions": ["[WALK] <bedroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'juice', 'microwave', 'toilet', 'walllamp', 'orchid', 'bathroom_counter', 'bookshelf', 'bookmark', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'wall_clock', 'food_food', 'coffe_maker', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'cupboard', 'bathtub', 'home_office', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'cat', 'cpuscreen', 'desk', 'bookshelf', 'computer', 'alcohol', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'stereo', 'shower', 'closetdrawer', 'candle', 'basket_for_clothes', 'photoframe', 'food_food', 'check', 'television', 'water', 'curtain', 'lighter', 'doorjamb', 'vacuum_cleaner', 'cupboard', 'home_office', 'table', 'bathroom_cabinet']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'nail_polish', 'detergent', 'orchid', 'drawing', 'hanger', 'floor', 'phone', 'desk', 'button', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'dining_room', 'nightstand', 'wall_clock', 'oven', 'cup', 'photoframe', 'bed', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go into the living room and turn on the radio.  Then, I select the station I want, which is usually a preset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[FIND] <button>', '[PUSH] <button>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'beer', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'mat', 'stereo', 'couch', 'pillow', 'bowl', 'wall', 'nightstand', 'powersocket', 'cup', 'toilet_paper', 'television', 'after_shave', 'rag', 'doorjamb', 'coffee_pot', 'dishrack', 'instrument_guitar', 'table', 'soap', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'toilet', 'chair', 'walllamp', 'floor', 'ceiling', 'phone', 'keyboard', 'envelope', 'toaster', 'ceilinglamp', 'stereo', 'faucet', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'clothes_gloves', 'novel', 'check', 'video_game_controller', 'television', 'kitchen_counter', 'hairdryer', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Walk to the sofa and sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'microwave', 'cd_player', 'freezer', 'band-aids', 'bathroom', 'food_carrot', 'orchid', 'music_stand', 'phone', 'bookshelf', 'thread', 'mail', 'stereo', 'homework', 'mousepad', 'closetdrawer', 'clothes_hat', 'form', 'broom', 'newspaper', 'photoframe', 'check', 'food_rice', 'food_donut', 'dresser', 'sink', 'napkin', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to the living and turn to the music stand. Switch on the CD player to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <music_stand>", "[FIND] <music_stand>", "[TURNTO] <music_stand>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'chair', 'walllamp', 'ceilingfan', 'knife', 'orchid', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'balanceball', 'mousepad', 'clothes_socks', 'bowl', 'wall', 'check', 'television', 'mouse', 'kitchen_counter', 'cd', 'door', 'light', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": When the phone rings, I pick up the call and give response to the call.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['juice', 'toilet', 'chair', 'freezer', 'detergent', 'food_carrot', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'bedroom', 'piano_bench', 'ceilinglamp', 'mat', 'shower', 'faucet', 'oil', 'food_orange', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'check', 'doorjamb', 'vacuum_cleaner', 'door', 'table', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Put the groceries on the table into the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTBACK] <food_chicken> <freezer>", "[CLOSE] <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_carrot> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'shower', 'window', 'clothes_hat', 'dining_room', 'wall', 'bench', 'powersocket', 'photoframe', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'table', 'light', 'tooth_paste']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Pick up phone,\n    \"actions\": ['[WALK] <table>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'food_ice_cream', 'microwave', 'chair', 'food_carrot', 'wallshelf', 'food_egg', 'cpuscreen', 'hanger', 'shampoo', 'phone', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'microphone', 'pillow', 'nightstand', 'bed', 'rag', 'cards', 'shoes', 'pajamas', 'door', 'sink', 'home_office', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the TV remote. I change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'computer', 'button', 'brush', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'standingmirror', 'coffe_maker', 'television', 'food_rice', 'wine', 'curtain', 'doorjamb', 'diary', 'pencil', 'door', 'bathroom_cabinet', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'food_ice_cream', 'hanger', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'faucet', 'ceilinglamp', 'wall', 'powersocket', 'photoframe', 'clothes_gloves', 'bed', 'food_food', 'check', 'ice', 'doorjamb', 'pencil', 'kitchen_counter', 'laundry_detergent', 'door', 'napkin', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on the light", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'food_pizza', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'cleaning_solution', 'mail', 'brush', 'toaster', 'tvstand', 'balanceball', 'window', 'food_noodles', 'mousepad', 'food_peanut_butter', 'clothes_hat', 'dining_room', 'pasta', 'bench', 'photoframe', 'oven_mitts', 'ice', 'rag', 'dresser', 'cupboard', 'food_steak', 'maindoor', 'table']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'ground_coffee', 'knifeblock', 'hanger', 'floor', 'ceiling', 'purse', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'wall', 'cup', 'razor', 'coin', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "To go kitchen and find a cup. open the faucet and fill water in the cup and drink.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <cup>", "[GRAB] <cup>", "[TURNTO] <faucet>", "[POINTAT] <faucet>", "[SWITCHOFF] <faucet>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'freezer', 'trashcan', 'orchid', 'drawing', 'bills', 'wallshelf', 'cpuscreen', 'knife', 'ceiling', 'character', 'stamp', 'keyboard', 'computer', 'ceilinglamp', 'couch', 'pillow', 'pasta', 'bench', 'oven', 'keys', 'novel', 'mouthwash', 'ice', 'tray', 'towel_rack', 'mouse', 'cupboard', 'bathtub']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the TV remote. I change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'chair', 'walllamp', 'detergent', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'table_cloth', 'mop_bucket', 'ceilinglamp', 'tvstand', 'mat', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'check', 'television', 'tray', 'wooden_spoon', 'kettle', 'controller', 'food_steak', 'sink']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room and sit on the couch, ease into the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'tablelamp', 'food_ice_cream', 'toilet', 'bathroom', 'toothbrush', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceilinglamp', 'faucet', 'clothes_socks', 'closetdrawer', 'tape', 'cup', 'check', 'food_food', 'razor', 'food_onion', 'centerpiece', 'mouse', 'dresser', 'bathtub', 'bathroom_cabinet', 'light', 'clothes_skirt', 'tooth_paste', 'shoe_rack']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": EARLY MORNING AND BEFORE GOING TO BED,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'walllamp', 'bathroom', 'orchid', 'floor', 'ceiling', 'remote_control', 'computer', 'tvstand', 'faucet', 'window', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'nightstand', 'wall', 'cup', 'check', 'mouthwash', 'television', 'coffee_filter', 'curtain', 'after_shave', 'dresser', 'hairbrush', 'sink', 'table', 'board_game']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into living room. Sit on couch. Pick up remote. Turn on TV. Watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'toilet', 'chair', 'walllamp', 'trashcan', 'food_carrot', 'clothes_underwear', 'hanger', 'longboard', 'floor', 'ceiling', 'toaster', 'electric_shaver', 'couch', 'wall', 'nightstand', 'oven', 'cup', 'coffe_maker', 'food_food', 'television', 'cards', 'mouse', 'kitchen_counter', 'maindoor', 'home_office', 'light', 'dishwasher', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Making coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['microwave', 'bathroom', 'cat', 'food_carrot', 'toothbrush', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'toaster', 'mat', 'faucet', 'closetdrawer', 'couch', 'form', 'wall', 'novel', 'coffe_maker', 'wine', 'kettle', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'light', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Washing Hands", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[TOUCH] <soap>", "[PUTOBJBACK] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'food_butter', 'knifeblock', 'floor', 'ceiling', 'character', 'desk', 'food_cake', 'computer', 'bookmark', 'toaster', 'stovefan', 'mat', 'faucet', 'food_kiwi', 'wall', 'oven', 'powersocket', 'food_vegetable', 'bed', 'food_food', 'oven_mitts', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I prepare meals and cook them on the stove.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[CLOSE] <freezer>', '[FIND] <oven>', '[SWITCHON] <oven>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'homework', 'foundation', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'bed', 'novel', 'check', 'television', 'doorjamb', 'mouse', 'door', 'sink', 'light', 'table', 'dishwasher']\nTask:Work\nresponse:\n", "output": {"thoughts": "Find a PC and do homework", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[FIND] <mouse>", "[FIND] <check>", "[FIND] <homework>", "[GRAB] <homework>", "[FIND] <homework>", "[READ] <homework>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'stovefan', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'tape', 'newspaper', 'bed', 'television', 'doorjamb', 'mouse', 'sink', 'home_office', 'table', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office. Sit on chair. Switch on computer. type in keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'jelly', 'freezer', 'walllamp', 'knifeblock', 'drawing', 'floor', 'ceiling', 'food_cake', 'stovefan', 'toaster', 'plate', 'mat', 'shower', 'food_kiwi', 'wall', 'nightstand', 'oven', 'powersocket', 'photoframe', 'food_food', 'tray', 'food_onion', 'doorjamb', 'kitchen_counter', 'sauce_pan', 'cupboard', 'door', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <spoon>", "[GRAB] <spoon>", "[FIND] <sauce_pan>", "[PUTBACK] <spoon> <sauce_pan>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <plate>", "[PUTBACK] <pot> <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'ground_coffee', 'trashcan', 'desk', 'bedroom', 'purse', 'mop_bucket', 'shower', 'picture', 'homework', 'mousepad', 'form', 'wall', 'oven', 'bed', 'keys', 'video_game_controller', 'wine', 'food_onion', 'after_shave', 'console', 'mouse', 'cd', 'comb', 'dresser', 'door', 'clothes_skirt', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find the cat; pet the cat by touching it,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'floor', 'character', 'bookshelf', 'computer', 'toaster', 'mat', 'plate', 'window', 'food_noodles', 'food_orange', 'wall', 'oven', 'photoframe', 'keys', 'bed', 'coffe_maker', 'toothbrush_holder', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'clothes_dress', 'napkin', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'alcohol', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'iron', 'faucet', 'mousepad', 'bowl', 'wall', 'nightstand', 'oven', 'mouse', 'cupboard', 'door', 'clothes_scarf', 'bathroom_cabinet', 'table', 'instrument_guitar']\nTask:Work\nresponse:\n{\n    \"thoughts\": For doing homework walk to home office. Walk to computer. Switch on Computer. Sit in a chair and do homework in computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'microphone', 'form', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'coffe_maker', 'television', 'curtain', 'vacuum_cleaner', 'dishrack', 'dresser', 'hairbrush', 'door', 'table', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'toilet', 'chair', 'walllamp', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'ceiling', 'bedroom', 'bookshelf', 'food_bacon', 'table_cloth', 'toaster', 'laptop', 'ceilinglamp', 'floor_lamp', 'mousepad', 'couch', 'clothes_pants', 'towel', 'dining_room', 'basket_for_clothes', 'check', 'pencil', 'chessboard', 'food_steak', 'napkin', 'spoon', 'table']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Walk to the lamp.  Find the switch.  Click the switch off.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[SWITCHOFF] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_pizza', 'filing_cabinet', 'toilet', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'folder', 'keyboard', 'faucet', 'iron', 'mousepad', 'pillow', 'dining_room', 'tape', 'cutting_board', 'cup', 'food_food', 'video_game_controller', 'food_rice', 'food_donut', 'rag', 'kettle', 'mouse', 'kitchen_counter', 'controller', 'food_steak', 'sink', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find a radio there and switch it on to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'jelly', 'walllamp', 'trashcan', 'hanger', 'longboard', 'floor', 'desk', 'mail', 'toaster', 'plate', 'shower', 'balanceball', 'couch', 'wall', 'oven', 'box', 'cup', 'bed', 'coffe_maker', 'tray', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'vase', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['food_salt', 'crayon', 'microwave', 'food_carrot', 'orchid', 'knifeblock', 'bills', 'floor', 'bedroom', 'stamp', 'envelope', 'clothes_jacket', 'bookmark', 'toaster', 'faucet', 'towel', 'dining_room', 'wall', 'coffe_maker', 'food_food', 'video_game_controller', 'tray', 'lighter', 'diary', 'kitchen_counter', 'cupboard', 'sink', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I have had my lunch and walked towards the sink and applied soap to my hands and washed cleanly.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'jelly', 'freezer', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'purse', 'computer', 'iron', 'window', 'mousepad', 'oil', 'wall', 'powersocket', 'bed', 'keys', 'mouthwash', 'wine', 'doorjamb', 'mouse', 'cupboard', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'food_bread', 'toilet', 'freezer', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'shampoo', 'bedroom', 'ceilinglamp', 'closetdrawer', 'form', 'towel', 'broom', 'nightstand', 'wall', 'keys', 'novel', 'coffe_maker', 'coin', 'water', 'curtain', 'headset', 'dishrack', 'hair', 'controller', 'sink', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, look at the water glass, grab that glass, find the water cane, open the water cane tap, drink water", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[TURNTO] <water_glass>", "[LOOKAT] <water_glass>", "[GRAB] <water_glass>", "[FIND] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'toilet', 'chair', 'blow_dryer', 'hanger', 'bathroom_counter', 'ceiling', 'phone', 'desk', 'mail', 'shoe-shine_kit', 'mousepad', 'food_noodles', 'cutting_board', 'cup', 'photoframe', 'bed', 'coffe_maker', 'television', 'curtain', 'centerpiece', 'comb', 'pajamas', 'clothes_dress', 'door', 'home_office', 'light', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into the living room, sit down by the cat, put out hand, stroke the cat from head to tail,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_ice_cream', 'toilet', 'freezer', 'drawing', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'bowl', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'bed', 'food_food', 'coffe_maker', 'wooden_spoon', 'doorjamb', 'pencil', 'comb', 'door', 'light', 'soap', 'scrabble', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "In the kitchen where I left the bags of groceries, I grab cheese from the grocery bag, I walk over to the refrigerator, open it up, neatly put cheese in the fridge, Close the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'shoe-shine_kit', 'balanceball', 'picture', 'ceilinglamp', 'mousepad', 'wall', 'nightstand', 'newspaper', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I need to finish some work on my computer in my home office to meet a deadline for tomorrow.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[TYPE] <keyboard>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'water_glass', 'food_pizza', 'juice', 'toilet', 'chair', 'ground_coffee', 'nail_polish', 'food_butter', 'wallshelf', 'cpuscreen', 'kitchen_cabinet', 'character', 'purse', 'piano_bench', 'couch', 'novel', 'coffe_maker', 'razor', 'water', 'tray', 'after_shave', 'chessboard', 'laundry_detergent', 'food_steak', 'door', 'napkin', 'light', 'table', 'creditcard']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a glass of water to drink", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'mousepad', 'couch', 'food_kiwi', 'wall', 'bench', 'cup', 'check', 'television', 'curtain', 'napkin', 'table', 'soap', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch. I pick up the remote control. I push the power button.  I push the guide button and look for my favorite show \"The Middle.\"  I click that channel and enjoy my show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUSH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'detergent', 'floor', 'ceiling', 'character', 'bookshelf', 'toaster', 'mat', 'deck_of_cards', 'food_noodles', 'closetdrawer', 'mousepad', 'form', 'towel', 'wall', 'pasta', 'powersocket', 'light_bulb', 'novel', 'video_game_controller', 'television', 'tray', 'doorjamb', 'centerpiece', 'controller', 'door', 'instrument_guitar', 'light', 'dishwasher', 'bathroom_cabinet']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go into my room, find the light,  then turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'cat', 'drawing', 'knife', 'knifeblock', 'hanger', 'floor', 'bedroom', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'faucet', 'iron', 'window', 'mousepad', 'couch', 'form', 'dining_room', 'wall', 'oven', 'box', 'newspaper', 'bed', 'novel', 'toilet_paper', 'television', 'tray', 'sink', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'water_glass', 'toilet', 'jelly', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'stovefan', 'mat', 'shower', 'food_kiwi', 'wall', 'nightstand', 'oven', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'water', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'door', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I need to go to Kitchen.  Get a cup.  Fill cup.  Drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'bowl', 'wall', 'powersocket', 'cup', 'novel', 'coffe_maker', 'food_donut', 'doorjamb', 'mouse', 'dresser', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find a desk and grab a book. Walk to the chair, sit in the chair and read a book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'band-aids', 'freezer', 'bathroom', 'detergent', 'food_butter', 'drawing', 'wallshelf', 'bookshelf', 'computer', 'tvstand', 'shower', 'homework', 'clothes_pants', 'bowl', 'nightstand', 'wall', 'oven', 'cup', 'photoframe', 'food_food', 'toilet_paper', 'coffe_maker', 'wine', 'tray', 'kitchen_counter', 'cupboard', 'door', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat.  Pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'cat', 'trashcan', 'drawing', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bookshelf', 'keyboard', 'mail', 'toaster', 'mat', 'tvstand', 'couch', 'pillow', 'broom', 'wall', 'nightstand', 'powersocket', 'television', 'food_donut', 'doorjamb', 'door', 'bathroom_cabinet', 'light', 'table', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": My cat looks like it wants attention, so I will go pet it on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'toilet', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'envelope', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'stereo', 'mousepad', 'towel', 'wall', 'powersocket', 'photoframe', 'bed', 'toilet_paper', 'tray', 'curtain', 'doorjamb', 'dresser', 'controller', 'door', 'food_steak', 'light', 'tooth_paste']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "walk to bathroom , find shower , take a shower", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'blow_dryer', 'nail_polish', 'ground_coffee', 'milk', 'knife', 'wallshelf', 'hanger', 'character', 'desk', 'keyboard', 'stovefan', 'mat', 'plate', 'food_jam', 'nightstand', 'wall', 'toy', 'bed', 'coffe_maker', 'wine', 'rag', 'centerpiece', 'pencil', 'dresser', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Pick up the book, open to current page, read the words.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'jelly', 'bathroom', 'detergent', 'dough', 'orchid', 'floor', 'bedroom', 'mat', 'dish_soap', 'mousepad', 'dining_room', 'wall', 'nightstand', 'oven', 'food_cheese', 'photoframe', 'coffe_maker', 'food_food', 'ice', 'tray', 'kitchen_counter', 'headset', 'dishrack', 'dresser', 'cupboard', 'door', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the dishwasher and put dish soap. I close the dishwasher and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'walllamp', 'drawing', 'orchid', 'floor', 'phone', 'character', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'wall_clock', 'bed', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'hairdryer', 'sink', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the sofa. Sit down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'chair', 'nail_polish', 'knife', 'cpuscreen', 'floor', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'mousepad', 'closetdrawer', 'couch', 'dining_room', 'wall', 'basket_for_clothes', 'cup', 'check', 'television', 'curtain', 'mouse', 'chessboard', 'shoes', 'comb', 'cupboard', 'hairdryer', 'door', 'home_office']\nTask:Read book\nresponse:\n", "output": {"thoughts": "find a book to read. sit down. read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'oil', 'couch', 'wall', 'cup', 'food_food', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'console', 'hairbrush', 'cupboard', 'food_steak', 'maindoor', 'door', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch favorite shows, netflix, hulu,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'walllamp', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'bookshelf', 'food_cake', 'plate', 'iron', 'dish_soap', 'mousepad', 'dining_room', 'wall', 'cup', 'newspaper', 'novel', 'coffe_maker', 'food_food', 'check', 'coin', 'oven_mitts', 'food_rice', 'curtain', 'mouse', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk into kitchen, find dishwasher, put plates in dishwasher, put soap in dishwasher, turn on dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'microwave', 'walllamp', 'toothbrush', 'drawing', 'knifeblock', 'wallshelf', 'keyboard', 'food_cake', 'ceilinglamp', 'tvstand', 'shower', 'stereo', 'oil', 'tooth_paste', 'couch', 'closetdrawer', 'dining_room', 'toy', 'basket_for_clothes', 'light_bulb', 'check', 'centerpiece', 'kettle', 'mouse', 'kitchen_counter', 'shoes', 'light', 'board_game']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": EARLY MORNING AND BEFORE GOING TO BED,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'sponge', 'microwave', 'toilet', 'freezer', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'bedroom', 'purse', 'mat', 'ceilinglamp', 'shower', 'picture', 'wall', 'powersocket', 'oven', 'toilet_paper', 'food_donut', 'tray', 'curtain', 'doorjamb', 'cupboard', 'door', 'hairdryer', 'spoon', 'light', 'dirt']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. I walk over to the toilet. I sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'tea_bag', 'bookshelf', 'ceilinglamp', 'dining_room', 'wall', 'oven', 'photoframe', 'bed', 'novel', 'check', 'tray', 'doorjamb', 'towel_rack', 'console', 'pencil', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the Living Room. Look at the Cabinet. Find a Textbook. Grab the textbook. Open the Textbook. Read the textbook.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[FIND] <filing_cabinet>', '[TURNTO] <filing_cabinet>', '[LOOKAT] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'tea', 'microwave', 'walllamp', 'detergent', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'stovefan', 'toaster', 'faucet', 'stereo', 'bowl', 'food_kiwi', 'wall', 'oven', 'bed', 'coffe_maker', 'food_food', 'drinking_glass', 'food_donut', 'tray', 'after_shave', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'clothes_skirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'crayon', 'microwave', 'nail_polish', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'couch', 'dining_room', 'wall', 'nightstand', 'oven', 'novel', 'television', 'curtain', 'dresser', 'pajamas', 'table', 'instrument_violin']\nTask:Turn off TV\nresponse:\n{\n    \"thoughts\": Walk over to tv and press power button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'blow_dryer', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'television', 'ice', 'rag', 'doorjamb', 'mouse', 'controller', 'door', 'light', 'table', 'scrabble', 'dirt']\nTask:Work\nresponse:\n", "output": {"thoughts": "walk to home office ,find computer ,switch on computer and work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'dog', 'toothbrush', 'bills', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'purse', 'faucet', 'balanceball', 'food_noodles', 'wall', 'bench', 'cup', 'photoframe', 'mouthwash', 'console', 'kitchen_counter', 'cupboard', 'sink', 'bag', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": go to bathroom. walk to sink,turn on faucet, put hands in water. rub hands together. grab soap. rub soap in hands. put soap down. put hands under water. rubs hands together. turn off faucet,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['spectacles', 'crayon', 'microwave', 'toilet', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'shampoo', 'computer', 'mat', 'tvstand', 'picture', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'bathroom_cabinet', 'light', 'instrument_violin']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone. .I reach my arm out. I open my hand. I place my hand on the phone. I close my fingers around the phone. I pick up the phone.", "actions": ["[WALK] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'dog', 'trashcan', 'drawing', 'cpuscreen', 'character', 'keyboard', 'bookshelf', 'toaster', 'mat', 'tvstand', 'faucet', 'window', 'mousepad', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'wooden_spoon', 'kettle', 'shaving_cream', 'towel_rack', 'cupboard', 'conditioner', 'bathtub', 'dirt']\nTask:Lock door\nresponse:\n{\n    \"thoughts\": Turn the lock.,\n    \"actions\": ['[WALK] <bedroom>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'orchid', 'wallshelf', 'clothes_underwear', 'cpuscreen', 'phone', 'character', 'bedroom', 'keyboard', 'toaster', 'mat', 'plate', 'picture', 'foundation', 'dish_soap', 'clothes_socks', 'pillow', 'powersocket', 'cutting_board', 'basket_for_clothes', 'bed', 'television', 'curtain', 'diary', 'cupboard', 'sink', 'spoon', 'home_office']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Pick up dish.  Put in soapy water.  Wash dish.", "actions": ["[WALK] <dining_room>", "[WALK] <dish_soap>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <dish_soap> <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[WIPE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'bathroom', 'knifeblock', 'wallshelf', 'bathroom_counter', 'character', 'bedroom', 'tea_bag', 'keyboard', 'food_cake', 'toaster', 'ceilinglamp', 'shower', 'faucet', 'couch', 'food_jam', 'clothes_pants', 'bowl', 'dining_room', 'nightstand', 'oven', 'ice', 'lighter', 'mouse', 'conditioner', 'door', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to the living room, Find Phone stand. Find Phone. Pickup the receiver to attend the call.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'keyboard', 'remote_control', 'bookshelf', 'computer', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'broom', 'nightstand', 'wall', 'television', 'curtain', 'doorjamb', 'pencil', 'dresser', 'pajamas', 'laundry_detergent', 'clothes_dress', 'bag', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk into the living room walk to the remote pick it up and turn in the tv and watch", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[PUTOBJBACK] <remote_control>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'food_salt', 'chair', 'cpuscreen', 'hanger', 'kitchen_cabinet', 'bathroom_counter', 'desk', 'bookshelf', 'thread', 'stovefan', 'laptop', 'faucet', 'window', 'clothes_socks', 'closetdrawer', 'powersocket', 'food_cheese', 'water', 'drinking_glass', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'dishrack', 'sink', 'table', 'colander']\nTask:Drink\nresponse:\n", "output": {"thoughts": "You walk to the kitchen and open the kitchen cabinet and grab a drinking glass. You find the water and pour the water into the drinking glass. You drink the water from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'food_pizza', 'tablelamp', 'microwave', 'floor', 'ceiling', 'desk', 'bookshelf', 'stovefan', 'mat', 'ceilinglamp', 'laptop', 'window', 'homework', 'oil', 'closetdrawer', 'couch', 'nightstand', 'wall', 'cup', 'bed', 'food_food', 'mouthwash', 'toothbrush_holder', 'oven_mitts', 'razor', 'food_donut', 'check', 'clothes_scarf', 'bathroom_cabinet']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Walk to bedroom. Walk to bed, lie in bed.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['tea', 'toilet', 'chair', 'freezer', 'ground_coffee', 'drawing', 'floor', 'phone', 'alcohol', 'mail', 'piano_bench', 'mat', 'shower', 'stereo', 'homework', 'food_peanut_butter', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'bowl', 'powersocket', 'curtain', 'doorjamb', 'cd', 'laundry_detergent', 'door', 'sink', 'light', 'soap']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to bathroom, I open the door of the bathroom, and sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'drawing', 'orchid', 'bills', 'floor', 'desk', 'mat', 'tvstand', 'stereo', 'clothes_socks', 'closetdrawer', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'basket_for_clothes', 'photoframe', 'newspaper', 'clothes_gloves', 'standingmirror', 'mouthwash', 'television', 'razor', 'towel_rack', 'kitchen_counter', 'clothes_dress', 'sink', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": put clothes and soap in the washing machine, turn it on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[FIND] <soap>', '[GRAB] <soap>', '[OPEN] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'mop', 'chair', 'walllamp', 'food_butter', 'food_egg', 'cpuscreen', 'floor', 'ceiling', 'stamp', 'mail', 'stovefan', 'ceilinglamp', 'stereo', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'television', 'curtain', 'after_shave', 'mouse', 'hair', 'food_steak', 'sink', 'light', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and I lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'toilet', 'bathroom_counter', 'floor', 'phone', 'desk', 'stovefan', 'toaster', 'plate', 'tvstand', 'picture', 'mat', 'mousepad', 'dining_room', 'food_kiwi', 'basket_for_clothes', 'cup', 'photoframe', 'bed', 'novel', 'keys', 'coin', 'food_rice', 'tray', 'pencil', 'kitchen_counter', 'door', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the cat. I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'microwave', 'walllamp', 'bathroom', 'orchid', 'drawing', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'toaster', 'tvstand', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'bed', 'keys', 'food_food', 'television', 'tray', 'curtain', 'kettle', 'dresser', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room. Sit on the couch. Turn on TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'crayon', 'toilet', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'mail', 'food_bacon', 'clothes_jacket', 'table_cloth', 'faucet', 'mousepad', 'closetdrawer', 'food_jam', 'nightstand', 'powersocket', 'novel', 'food_food', 'centerpiece', 'diary', 'console', 'cupboard', 'bathroom_cabinet', 'table', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Pick up the ringing phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[POINTAT] <table>', '[FIND] <phone>', '[GRAB] <phone>', '[PUTOBJBACK] <phone>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'nail_polish', 'orchid', 'bills', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'doorjamb', 'mouse', 'door', 'home_office', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "walk into Living room, look for computer, walk to computer, turn on computer, grab mouse, click mouse", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[WALK] <computer>", "[TURNTO] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'bathroom', 'detergent', 'nail_polish', 'trashcan', 'food_butter', 'orchid', 'hanger', 'longboard', 'floor', 'bathroom_counter', 'remote_control', 'envelope', 'computer', 'window', 'electric_shaver', 'wall', 'cutting_board', 'bed', 'television', 'tray', 'curtain', 'after_shave', 'lighter', 'mouse', 'cupboard', 'clothes_dress', 'light', 'clothes_skirt', 'board_game']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watching tv with a remote is always fun; lay back, relax and enjoy;,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['pot', 'tea', 'microwave', 'band-aids', 'walllamp', 'detergent', 'knifeblock', 'floor', 'ceiling', 'shampoo', 'stovefan', 'toaster', 'faucet', 'dining_room', 'bowl', 'wall', 'oven', 'powersocket', 'cup', 'coffe_maker', 'check', 'oven_mitts', 'water', 'tray', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to the kitchen, grab a cup from the cupboard, go to the sink, put cup under faucet, turn on faucet til cup is full, turn off faucet, drink from cup", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[PUTBACK] <cup> <sink>", "[SWITCHON] <faucet>", "[FIND] <water>", "[GRAB] <cup>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'crayon', 'band-aids', 'nail_polish', 'cat', 'food_carrot', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'food_bacon', 'ceilinglamp', 'shoe-shine_kit', 'window', 'couch', 'food_kiwi', 'wall', 'bed', 'keys', 'check', 'television', 'oven_mitts', 'curtain', 'conditioner', 'maindoor', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": The cat is laying on the couch in the living room. I want to pet it, so I am going to walk to the living room, sit on the couch, and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'knifeblock', 'drawing', 'food_egg', 'desk', 'keyboard', 'food_cake', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'brush', 'window', 'foundation', 'deck_of_cards', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'light_bulb', 'clothes_gloves', 'check', 'mouthwash', 'curtain', 'dresser', 'clothes_dress', 'sink', 'creditcard']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the radio.  I switch on the radio and listen", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'keyboard', 'thread', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'mousepad', 'couch', 'wall', 'photoframe', 'bed', 'toilet_paper', 'food_onion', 'towel_rack', 'mouse', 'kitchen_counter', 'cd', 'hairbrush', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'freezer', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'shampoo', 'stovefan', 'mat', 'shower', 'window', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'newspaper', 'food_food', 'drinking_glass', 'water', 'curtain', 'doorjamb', 'dresser', 'food_steak', 'door', 'home_office', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tea', 'jelly', 'dog', 'chair', 'ground_coffee', 'bathroom', 'phone', 'bedroom', 'stovefan', 'mat', 'plate', 'faucet', 'shoe-shine_kit', 'wall', 'nightstand', 'powersocket', 'cup', 'basket_for_clothes', 'newspaper', 'food_food', 'coffe_maker', 'tray', 'curtain', 'food_steak', 'door', 'bag', 'instrument_guitar', 'light', 'dishwasher']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <home_office>', '[DROP] <plate>', '[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <home_office>', '[DROP] <cup>', '[WALK] <dining_room>', '[WALK] <fork>', '[FIND] <fork>', '[GRAB] <fork>', '[WALK] <home_office>', '[DROP] <fork>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'orchid', 'needle', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to my desk. I turn on the computer. I open a web page and look at it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'beer', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'food_cake', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'photoframe', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'door', 'bathtub', 'bathroom_cabinet', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Sit on sofa. Lie down on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'bathroom', 'food_carrot', 'wallshelf', 'ceiling', 'dvd_player', 'character', 'bedroom', 'desk', 'keyboard', 'thread', 'stovefan', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'cup', 'bed', 'food_food', 'razor', 'toothbrush_holder', 'curtain', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'napkin', 'table', 'clothes_skirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Go to the living room.  Find the telephone.  Lift up the receiver.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'juice', 'microwave', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'shower', 'faucet', 'window', 'mousepad', 'food_orange', 'wall', 'television', 'curtain', 'after_shave', 'mouse', 'dresser', 'door', 'napkin', 'table']\nTask:Clean screen\nresponse:\n{\n    \"thoughts\": Use a soft rag to wipe the screen until there are no marks on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <computer>', '[WIPE] <computer>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bedroom', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'food_food', 'television', 'wooden_spoon', 'curtain', 'diary', 'dresser', 'cupboard', 'clothes_dress', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "You walk to the living room and sit on the sofa. You grab the remote control, look at it and point at the television, you watch television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'microwave', 'toilet', 'walllamp', 'knifeblock', 'wallshelf', 'floor', 'shampoo', 'character', 'computer', 'bookmark', 'toaster', 'plate', 'faucet', 'shoe-shine_kit', 'dish_soap', 'wall', 'toy', 'light_bulb', 'coffe_maker', 'rag', 'kitchen_counter', 'hairbrush', 'cupboard', 'dresser', 'controller', 'sink', 'home_office', 'table', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Make some dish water, grab a dish rag, put dishes in dish water and wash.,\n    \"actions\": ['[RUN] <sink>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <plate>', '[PUTBACK] <dish_soap> <plate>', '[RINSE] <dish_soap>', '[FIND] <plate>', '[WASH] <plate>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'walllamp', 'knifeblock', 'orchid', 'floor', 'phone', 'ceiling', 'bedroom', 'keyboard', 'food_bacon', 'stovefan', 'toaster', 'plate', 'faucet', 'window', 'homework', 'closetdrawer', 'wall', 'oven', 'ice', 'tray', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'hairdryer', 'sink', 'bathroom_cabinet', 'coffee']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[TURNTO] <plate>", "[LOOKAT] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[WASH] <plate>", "[PUTBACK] <plate> <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'chair', 'bathroom', 'food_butter', 'phone', 'ceiling', 'bedroom', 'tea_bag', 'laptop', 'ceilinglamp', 'shower', 'iron', 'clothes_socks', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'powersocket', 'standingmirror', 'photoframe', 'toothbrush_holder', 'food_donut', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'controller', 'door', 'table']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'orchid', 'bills', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'bed', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'comb', 'door', 'home_office', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to the living room, where my computer is. I locate it, switch it on and type desired URL adress in the web browser.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'detergent', 'orchid', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'computer', 'tvstand', 'faucet', 'plate', 'window', 'oil', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'bed', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'juice', 'tea', 'toilet', 'freezer', 'walllamp', 'dough', 'drawing', 'food_egg', 'floor', 'dvd_player', 'character', 'alcohol', 'stovefan', 'mat', 'shower', 'closetdrawer', 'clothes_hat', 'food_jam', 'wall', 'powersocket', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'laundry_detergent', 'door', 'light', 'dishwasher', 'table']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[TURNTO] <toilet>", "[LOOKAT] <toilet>", "[WALK] <toilet>", "[TURNTO] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'ceilingfan', 'trashcan', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bookshelf', 'table_cloth', 'mop_bucket', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'food_noodles', 'couch', 'wall', 'nightstand', 'food_food', 'television', 'curtain', 'cupboard', 'table', 'light', 'vase', 'dishwasher', 'dirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch or the chair or the floor and watch cable, movies, or netflix.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'filing_cabinet', 'food_bread', 'toilet', 'jelly', 'freezer', 'food_butter', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'brush', 'mat', 'shower', 'mousepad', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'television', 'ice', 'doorjamb', 'hair', 'door', 'napkin', 'light', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen and put groceries in fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'tvstand', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'check', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'clothes_dress', 'door', 'conditioner', 'sink', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": My cat looks like it wants attention, so I will go pet it on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'microwave', 'detergent', 'knifeblock', 'floor', 'ceiling', 'character', 'computer', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'wall', 'cup', 'check', 'mouthwash', 'water', 'wine', 'after_shave', 'curtain', 'food_onion', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'conditioner', 'sink', 'napkin', 'bathroom_cabinet']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <cup>", "[GRAB] <cup>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'scissors', 'chair', 'bathroom', 'detergent', 'cat', 'ceilingfan', 'dough', 'knifeblock', 'hanger', 'floor', 'ceiling', 'character', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'microphone', 'wall', 'oven', 'coffe_maker', 'tray', 'curtain', 'lighter', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find a cat in the kitchen and touch it,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['spectacles', 'chair', 'walllamp', 'orchid', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'desk', 'bedroom', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'iron', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'bed', 'television', 'doorjamb', 'mouse', 'door', 'home_office', 'light', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "enter office, find computer, turn it on,", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TOUCH] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'tea_bag', 'computer', 'mat', 'tvstand', 'mousepad', 'couch', 'candle', 'pillow', 'bowl', 'wall', 'oven', 'toy', 'cup', 'photoframe', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'dresser', 'laundry_detergent', 'sink', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, find the sofa and sit in the sofa and find pillow. lie down in sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <pillow>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'detergent', 'bills', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'mat', 'laptop', 'shower', 'faucet', 'ceilinglamp', 'mousepad', 'microphone', 'form', 'pillow', 'broom', 'wall', 'newspaper', 'after_shave', 'doorjamb', 'cd', 'comb', 'controller', 'door', 'bathroom_cabinet', 'light', 'table', 'creditcard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.", "actions": ["[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'walllamp', 'ceilingfan', 'orchid', 'hanger', 'longboard', 'floor', 'ceiling', 'character', 'bedroom', 'folder', 'bookshelf', 'toaster', 'ceilinglamp', 'plate', 'dish_soap', 'candle', 'wall', 'oven', 'bench', 'box', 'cup', 'clothes_gloves', 'coffe_maker', 'tray', 'console', 'maindoor', 'door', 'napkin', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['scissors', 'dog', 'walllamp', 'clothes_underwear', 'bathroom_counter', 'phone', 'character', 'desk', 'tea_bag', 'envelope', 'toaster', 'ceilinglamp', 'window', 'mousepad', 'food_kiwi', 'nightstand', 'cutting_board', 'photoframe', 'coffe_maker', 'television', 'tray', 'rag', 'doorjamb', 'diary', 'cupboard', 'hairdryer', 'home_office', 'light', 'table', 'creditcard']\nTask:Receive credit card\nresponse:\n", "output": {"thoughts": "hear the dooble, go to door, take de envelope, open it with a scissor, take the creditcard", "actions": ["[WALK] <home_office>", "[WALK] <envelope>", "[FIND] <envelope>", "[GRAB] <envelope>", "[WALK] <home_office>", "[WALK] <scissors>", "[FIND] <scissors>", "[OPEN] <envelope>", "[FIND] <creditcard>", "[GRAB] <creditcard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'sheets', 'microwave', 'toilet', 'freezer', 'trashcan', 'food_carrot', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'ceilinglamp', 'shower', 'oil', 'pillow', 'wall', 'novel', 'food_food', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'home_office', 'light']\nTask:Put away groceries\nresponse:\n{\n    \"thoughts\": Place groceries in refrigerator from shopping at the grocery store.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'crayon', 'microwave', 'freezer', 'detergent', 'drawing', 'knife', 'cpuscreen', 'purse', 'stovefan', 'brush', 'mat', 'faucet', 'food_noodles', 'couch', 'dustpan', 'nightstand', 'wall', 'powersocket', 'newspaper', 'novel', 'ice', 'kitchen_counter', 'hairbrush', 'cupboard', 'sink', 'table', 'clothes_skirt', 'board_game']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into my bedroom, go to the lamp and use the switch to turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[LOOKAT] <floor_lamp>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'ceilingfan', 'orchid', 'bathroom_counter', 'bookshelf', 'purse', 'stovefan', 'mat', 'iron', 'electric_shaver', 'food_noodles', 'mousepad', 'couch', 'clothes_pants', 'pillow', 'powersocket', 'keys', 'coffe_maker', 'oven_mitts', 'wine', 'shaving_cream', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'home_office', 'light', 'tooth_paste']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'scissors', 'bathroom', 'drawing', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'laptop', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'wall', 'powersocket', 'photoframe', 'television', 'doorjamb', 'lighter', 'mouse', 'comb', 'door', 'sink', 'light', 'dishwasher', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Switch on light in home office", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_bread', 'bathroom', 'trashcan', 'food_carrot', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'phone', 'folder', 'bedroom', 'desk', 'keyboard', 'toaster', 'bench', 'tray', 'doorjamb', 'diary', 'console', 'mouse', 'towel_rack', 'cd', 'dresser', 'cupboard', 'controller', 'conditioner', 'table', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to table, grab water glass and drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>', '[PUTBACK] <water_glass> <table>']\n}\n", "input": "\nScene_Objects:['beer', 'microwave', 'dog', 'chair', 'cat', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'phone', 'character', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'closetdrawer', 'oven', 'food_cheese', 'novel', 'toothbrush_holder', 'television', 'food_donut', 'console', 'controller', 'clothes_scarf', 'light', 'colander', 'board_game', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my cat, and I pet him", "actions": ["[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'walllamp', 'bathroom', 'knifeblock', 'food_egg', 'food_bacon', 'brush', 'mat', 'deck_of_cards', 'balanceball', 'tvstand', 'ceilinglamp', 'picture', 'couch', 'wall', 'tape', 'cup', 'toilet_paper', 'food_food', 'video_game_controller', 'water', 'cupboard', 'bathroom_cabinet', 'table', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'spectacles', 'sheets', 'tablelamp', 'microwave', 'chair', 'jelly', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'stovefan', 'ceilinglamp', 'picture', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'wall_clock', 'photoframe', 'food_food', 'coffe_maker', 'television', 'centerpiece', 'coffee_pot', 'pencil', 'shoes', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Find the sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'dry_pasta', 'walllamp', 'trashcan', 'food_carrot', 'dough', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'wall', 'wall_clock', 'oven', 'food_cheese', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher']\nTask:Keep an eye on stove as something is cooking\nresponse:\n{\n    \"thoughts\": Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <pot>', '[TURNTO] <pot>', '[LOOKAT] <pot>', '[FIND] <wall_clock>', '[TURNTO] <wall_clock>', '[LOOKAT] <wall_clock>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'toilet', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'bowl', 'wall', 'television', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'light', 'table', 'board_game']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Turn on PC and log onto email account. Type email and send it. Switch off computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'chair', 'freezer', 'drawing', 'knifeblock', 'cpuscreen', 'needle', 'floor', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'brush', 'mousepad', 'wall', 'nightstand', 'bed', 'photoframe', 'check', 'food_food', 'doorjamb', 'mouse', 'kitchen_counter', 'headset', 'door', 'table', 'light', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I'll walk to my bedroom, where I have my computer. I'll walk to my computer and sit down in front of it. I'll put on my headphones and unlock my computer. I'll then navigate to music stored locally or streamed through the web.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <headset>', '[GRAB] <headset>', '[PUTON] <headset>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'detergent', 'orchid', 'knifeblock', 'knife', 'hanger', 'floor', 'character', 'bedroom', 'mop_bucket', 'toaster', 'stovefan', 'faucet', 'electric_shaver', 'couch', 'bowl', 'wall', 'box', 'food_cheese', 'coffe_maker', 'food_food', 'razor', 'shaving_cream', 'kitchen_counter', 'cupboard', 'sink', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen. Take the dishes from the sink. Apply dish washer, scrub and then wash with water.", "actions": ["[WALK] <sink>", "[FIND] <bowl>", "[GRAB] <bowl>", "[FIND] <dishwasher>", "[SCRUB] <dishwasher>", "[RINSE] <bowl>", "[OPEN] <dishwasher>", "[WASH] <bowl>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'laptop', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watching TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'detergent', 'orchid', 'clothes_underwear', 'floor', 'character', 'button', 'stovefan', 'ceilinglamp', 'toaster', 'homework', 'mousepad', 'couch', 'dustpan', 'microphone', 'form', 'towel', 'powersocket', 'basket_for_clothes', 'bed', 'light_bulb', 'oven_mitts', 'tray', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'colander', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to laundry room. Find laundry basket. Grab laundry basket. Find washing machine. Walk to washing machine. Open washing machine. Find detergent. Grab detergent. Put in detergent. Put in clothes. Close washing machine. Turn on machine.", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <washing_machine>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[PUTOBJBACK] <laundry_detergent>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'beer', 'microwave', 'chair', 'drawing', 'hanger', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'bookmark', 'laptop', 'ceilinglamp', 'faucet', 'shower', 'mousepad', 'couch', 'food_peanut_butter', 'photoframe', 'newspaper', 'razor', 'tray', 'kitchen_counter', 'vacuum_cleaner', 'clothes_dress', 'door', 'bathroom_cabinet', 'tooth_paste']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I go to my bed room and sit on the chair and in front of me is my desktop computer and start writing an email.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'keyboard', 'mat', 'ceilinglamp', 'shower', 'oil', 'food_peanut_butter', 'food_orange', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'pencil', 'dresser', 'door', 'napkin', 'light', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "put away groceries in fridge. open the door, put on shelf, close the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'ground_coffee', 'bathroom', 'shampoo', 'character', 'bedroom', 'bookshelf', 'keyboard', 'envelope', 'bookmark', 'shower', 'iron', 'foundation', 'mousepad', 'closetdrawer', 'couch', 'wall', 'oven', 'basket_for_clothes', 'cup', 'bed', 'food_food', 'curtain', 'towel_rack', 'cupboard', 'door', 'food_steak', 'bathroom_cabinet', 'dishwasher', 'dirt']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": go to the bathroom and brush your teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>']\n}\n", "input": "\nScene_Objects:['food_bread', 'tablelamp', 'chair', 'walllamp', 'floor', 'ceiling', 'bookshelf', 'mail', 'food_bacon', 'ceilinglamp', 'shower', 'couch', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'newspaper', 'coffe_maker', 'check', 'television', 'doorjamb', 'coffee_pot', 'vacuum_cleaner', 'shoes', 'dresser', 'pajamas', 'napkin', 'home_office', 'bathroom_cabinet', 'board_game']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Relax on the sofa in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'laptop', 'mat', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'cup', 'photoframe', 'television', 'doorjamb', 'mouse', 'door', 'creditcard']\nTask:Play on laptop\nresponse:\n{\n    \"thoughts\": Sit down on the couch. Place laptop on lap. Open laptop. Turn on laptop. Choose selected program. Play selected program. Close game. Close laptop.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <couch>', '[SIT] <couch>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <laptop>', '[PUTOBJBACK] <laptop>']\n}\n", "input": "\nScene_Objects:['glue', 'sponge', 'pot', 'filing_cabinet', 'dog', 'knife', 'wallshelf', 'hanger', 'ceiling', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'shower', 'closetdrawer', 'pillow', 'box', 'novel', 'food_onion', 'shaving_cream', 'pencil', 'chessboard', 'comb', 'dresser', 'clothes_scarf', 'dishwasher', 'board_game']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the home office. Pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'stereo', 'tvstand', 'mousepad', 'couch', 'broom', 'wall', 'bed', 'check', 'toilet_paper', 'television', 'ice', 'kettle', 'mouse', 'door', 'sink', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go in my office, sit at my desk turn on the computer, then type on my keyboard and open my web browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'walllamp', 'nail_polish', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'bedroom', 'keyboard', 'remote_control', 'button', 'bookmark', 'toaster', 'mat', 'faucet', 'food_jam', 'nightstand', 'tape', 'coffee_table', 'novel', 'check', 'mouse', 'pencil', 'chessboard', 'dresser', 'door', 'food_steak', 'laser_pointer']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'wall', 'powersocket', 'food_food', 'television', 'curtain', 'doorjamb', 'mouse', 'cupboard', 'door', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Turn on computer, pull up browser, browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'detergent', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'mail', 'toaster', 'stovefan', 'faucet', 'closetdrawer', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'oven', 'photoframe', 'television', 'coin', 'drinking_glass', 'water', 'curtain', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'sink', 'dishwasher', 'coffee']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and approach the sink. I take a drinking glass and set it under the faucet. I turn the faucet until the water fills the glass, then turn it back. Finally I lift the glass out of the sink and drink the water.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[GRAB] <drinking_glass>", "[FIND] <water>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'filing_cabinet', 'beer', 'measuring_cup', 'chair', 'ground_coffee', 'knifeblock', 'longboard', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'mail', 'toaster', 'picture', 'dustpan', 'wall', 'bench', 'cutting_board', 'tape', 'cup', 'coffe_maker', 'coffee_filter', 'water', 'after_shave', 'dresser', 'maindoor', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[GRAB] <ground_coffee>', '[PUTOBJBACK] <ground_coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sponge', 'scissors', 'microwave', 'bathroom', 'drawing', 'knifeblock', 'floor', 'bedroom', 'desk', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'closetdrawer', 'wall', 'powersocket', 'newspaper', 'photoframe', 'bed', 'check', 'food_rice', 'ice', 'curtain', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'instrument_guitar', 'board_game']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "go into the kitchen, walk to the sink. Turn on the water and rinse of a plate and clean them off fully with a sponge.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <plate>", "[RINSE] <plate>", "[FIND] <sponge>", "[GRAB] <sponge>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'desk', 'bookshelf', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'shower', 'electric_shaver', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'bed', 'television', 'oven_mitts', 'curtain', 'cupboard', 'bathtub', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Read magazine\nresponse:\n{\n    \"thoughts\": Please sit on the couch and read a magazine,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <newspaper>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <newspaper>', '[LOOKAT] <newspaper>', '[READ] <newspaper>', '[READ] <newspaper>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'blender', 'crayon', 'microwave', 'cat', 'drawing', 'orchid', 'needle', 'phone', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'food_jam', 'dining_room', 'powersocket', 'bed', 'photoframe', 'coffe_maker', 'television', 'diary', 'shoes', 'comb', 'controller', 'food_steak', 'instrument_violin', 'sink', 'dishwasher', 'board_game']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the cat. I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_salt', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'oil', 'wall', 'nightstand', 'powersocket', 'toy', 'bed', 'coffe_maker', 'television', 'doorjamb', 'centerpiece', 'mouse', 'dresser', 'hair', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, sit at desk, turn on computer, enter password, open application and begin work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'thread', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink', 'light', 'table', 'creditcard']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to home office. Walk to computer. Switch on Computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'deck_of_cards', 'faucet', 'mousepad', 'pillow', 'wall', 'oven', 'box', 'cup', 'toilet_paper', 'rag', 'kitchen_counter', 'cupboard', 'sink', 'bag', 'bathroom_cabinet', 'bathtub', 'table', 'home_office']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk the kitchen and walk up to the kitchen cabinet. I open the cabinet and take out a cup. I walk to sink and I turn the faucet on using the cold water handle. I put my cup under the running water until it is full. I turn off the cold water handle and bring the cup to my mouth and take a drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[WALK] <sink>', '[FIND] <faucet>', '[PUSH] <faucet>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'beer', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'stovefan', 'mat', 'toaster', 'shower', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'food_food', 'food_donut', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into the kitchen. Open the fridge. Take out the chicken. Open the kitchen cabinet. Take out the frying pan. Put the chicken in the frying pan. Put the frying pan on the stove. Turn on the stove. Cook the chicken.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'food_bread', 'chair', 'freezer', 'band-aids', 'walllamp', 'food_butter', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'button', 'clothes_jacket', 'toaster', 'ceilinglamp', 'microphone', 'wall', 'oven', 'powersocket', 'bed', 'food_food', 'coffe_maker', 'oven_mitts', 'ice', 'tray', 'curtain', 'pencil', 'clothes_dress', 'dishwasher']\nTask:Make toast\nresponse:\n{\n    \"thoughts\": Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[CLOSE] <freezer>', '[FIND] <toaster>', '[PLUGIN] <toaster>', '[PUTBACK] <food_bread> <toaster>', '[SWITCHON] <toaster>']\n}\n", "input": "\nScene_Objects:['glue', 'juice', 'toilet', 'freezer', 'knifeblock', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'mat', 'shower', 'dining_room', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'newspaper', 'food_food', 'curtain', 'doorjamb', 'cupboard', 'door', 'napkin', 'home_office', 'light', 'instrument_guitar']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_bread', 'filing_cabinet', 'tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'mousepad', 'food_noodles', 'pillow', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'doorjamb', 'towel_rack', 'mouse', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to home office ,find computer ,switch on computer and work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'mop', 'toilet', 'chair', 'freezer', 'cpuscreen', 'food_egg', 'floor', 'table_cloth', 'toaster', 'mat', 'plate', 'ceilinglamp', 'dish_soap', 'couch', 'food_peanut_butter', 'dustpan', 'bowl', 'wall', 'photoframe', 'coffe_maker', 'food_food', 'television', 'tray', 'kitchen_counter', 'cupboard', 'light', 'dishwasher', 'table']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Using dishwasher to wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'tablelamp', 'chair', 'drawing', 'cpuscreen', 'floor', 'phone', 'character', 'desk', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'tvstand', 'mousepad', 'form', 'pillow', 'wall', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'cards', 'door', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter office. Go to desk. Pick up phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['food_bread', 'food_pizza', 'tablelamp', 'microwave', 'floor', 'ceiling', 'desk', 'bookshelf', 'stovefan', 'mat', 'ceilinglamp', 'laptop', 'window', 'homework', 'oil', 'closetdrawer', 'couch', 'nightstand', 'wall', 'cup', 'bed', 'food_food', 'mouthwash', 'toothbrush_holder', 'oven_mitts', 'razor', 'food_donut', 'check', 'clothes_scarf', 'bathroom_cabinet']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Walk to bedroom. Walk to bed, lie in bed.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'bathroom', 'drawing', 'floor', 'character', 'bedroom', 'stamp', 'computer', 'mat', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'closetdrawer', 'clothes_hat', 'pillow', 'dining_room', 'broom', 'wall', 'bed', 'doorjamb', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'home_office', 'board_game', 'dirt']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to living room ,find electrical outlet , switch on eletrical outlet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <electrical_outlet>', '[FIND] <electrical_outlet>', '[SWITCHON] <electrical_outlet>']\n}\n", "input": "\nScene_Objects:['pot', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'stovefan', 'plate', 'shower', 'picture', 'faucet', 'dish_soap', 'pillow', 'wall', 'powersocket', 'cutting_board', 'tape', 'cup', 'coffe_maker', 'food_food', 'television', 'lighter', 'centerpiece', 'mouse', 'kitchen_counter', 'chessboard', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'chair', 'detergent', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'shower', 'mousepad', 'dustpan', 'dining_room', 'wall', 'powersocket', 'oven', 'check', 'television', 'doorjamb', 'hairbrush', 'controller', 'sink', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk to living room.  Find light switch.  Switch on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tea', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'cutting_board', 'cup', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'controller', 'door', 'light', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[PUSH] <mouse>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'phone', 'piano_bench', 'toaster', 'mat', 'faucet', 'shower', 'window', 'dining_room', 'wall', 'bench', 'hands_both', 'bed', 'television', 'oven_mitts', 'water', 'curtain', 'paper_towel', 'hairdryer', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. I turn on the water. I put soap on my hands. I rinse my hands. I turn off the water. I dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <water>', '[TURNTO] <water>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <water>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>']\n}\n", "input": "\nScene_Objects:['pot', 'measuring_cup', 'toilet', 'ground_coffee', 'knife', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'button', 'ceilinglamp', 'stereo', 'faucet', 'oil', 'pillow', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'coffee_filter', 'coin', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher', 'table', 'home_office']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open the coffee maker and put a filer, grab ground coffee and put it in the coffee maker, then close the coffee maker and switch it on", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'crayon', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'food_cake', 'ceilinglamp', 'tvstand', 'mat', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'food_food', 'television', 'curtain', 'pencil', 'kitchen_counter', 'door']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'ground_coffee', 'knifeblock', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'tea_bag', 'table_cloth', 'toaster', 'shoe-shine_kit', 'faucet', 'shower', 'window', 'homework', 'wall', 'basket_for_clothes', 'coffe_maker', 'water', 'drinking_glass', 'curtain', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'comb', 'cupboard', 'sink', 'light', 'clothes_skirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to kitchen, open cabinet, get glass, close cabinet, put glass under faucet, turn on water, turn off water when glass has 85% water, lift glass to mouth, open mouth and put on glass, tilt glass up and swallow water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[TURNTO] <kitchen_cabinet>", "[LOOKAT] <kitchen_cabinet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[FIND] <sink>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[SWITCHOFF] <faucet>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'mop', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'clothes_jacket', 'stovefan', 'faucet', 'window', 'pillow', 'bowl', 'wall', 'wall_clock', 'oven', 'cup', 'photoframe', 'food_food', 'toilet_paper', 'tray', 'after_shave', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'table', 'tooth_paste']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I switch on the stove and cook my groceries there.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <oven>', '[SWITCHON] <oven>', '[FIND] <bowl>', '[GRAB] <bowl>', '[PUTBACK] <bowl> <oven>', '[FIND] <food_food>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['tea', 'microwave', 'band-aids', 'bathroom', 'detergent', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'table_cloth', 'toaster', 'mat', 'tvstand', 'stereo', 'food_noodles', 'mousepad', 'food_jam', 'bowl', 'nightstand', 'wall', 'cup', 'bed', 'clothes_gloves', 'newspaper', 'coffe_maker', 'doorjamb', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into my bedroom and see my stereo. I decide to play music. I walk over to the stereo and turn it on. I sit down on the bed to listen to the music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'toilet', 'chair', 'freezer', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'character', 'bookshelf', 'mat', 'window', 'couch', 'dustpan', 'pillow', 'towel', 'wall', 'bench', 'food_food', 'television', 'coin', 'tray', 'curtain', 'doorjamb', 'bathtub', 'spoon', 'table', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dvd_player>', '[FIND] <dvd_player>', '[TURNTO] <dvd_player>', '[LOOKAT] <dvd_player>', '[SWITCHON] <dvd_player>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['blender', 'microwave', 'toilet', 'toothbrush', 'bills', 'orchid', 'character', 'bedroom', 'stamp', 'stovefan', 'laptop', 'ceilinglamp', 'shower', 'mat', 'faucet', 'towel', 'pillow', 'dining_room', 'oven', 'coffe_maker', 'television', 'toothbrush_holder', 'comb', 'cupboard', 'door', 'sink', 'table', 'soap', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Find toothbrush.  Find toothpaste.  Put toothpaste on toothbrush.  Brush teeth.", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>", "[FIND] <teeth>", "[SCRUB] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'food_ice_cream', 'blow_dryer', 'detergent', 'ceilingfan', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'shoe-shine_kit', 'couch', 'food_orange', 'wall', 'tape', 'television', 'curtain', 'kitchen_counter', 'conditioner', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": For relaxing walk to living room. Find the sofa to relax. Arrange the sofa to relax. Sit on the sofa to relaxing.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <conditioner>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['spectacles', 'chair', 'bathroom', 'food_carrot', 'knifeblock', 'floor', 'phone', 'bedroom', 'envelope', 'mail', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'tvstand', 'shower', 'window', 'candle', 'microphone', 'pillow', 'basket_for_clothes', 'toilet_paper', 'check', 'rag', 'chessboard', 'laundry_detergent', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Pick up my dirty clothes, load the washer, add detergent, turn on washer", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[RUN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'filing_cabinet', 'scissors', 'chair', 'ceilingfan', 'dough', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'coffee_table', 'television', 'coin', 'wooden_spoon', 'curtain', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'detergent', 'food_carrot', 'dough', 'drawing', 'character', 'bookshelf', 'keyboard', 'mail', 'bookmark', 'stovefan', 'iron', 'window', 'pillow', 'toy', 'cup', 'photoframe', 'food_food', 'check', 'television', 'toothbrush_holder', 'wooden_spoon', 'curtain', 'chessboard', 'hairbrush', 'dresser', 'sink', 'table', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find the cat, go near the cat and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'dresser', 'hairbrush', 'clothes_dress', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'freezer', 'detergent', 'food_carrot', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'desk', 'bookshelf', 'food_cake', 'keyboard', 'mail', 'mop_bucket', 'mat', 'faucet', 'window', 'food_noodles', 'mousepad', 'couch', 'cutting_board', 'food_food', 'shoes', 'dresser', 'sink', 'light', 'board_game', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go to the living room where the cat sleeps. Find cat. Proceed to petting cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'freezer', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'brush', 'tvstand', 'stereo', 'window', 'closetdrawer', 'towel', 'wall', 'powersocket', 'tape', 'food_food', 'television', 'curtain', 'doorjamb', 'vacuum_cleaner', 'dresser', 'door', 'clothes_scarf', 'dishwasher', 'board_game']\nTask:vacuum carpet\nresponse:\n{\n    \"thoughts\": I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'chair', 'walllamp', 'bathroom', 'drawing', 'knifeblock', 'floor', 'ceiling', 'thread', 'computer', 'ceilinglamp', 'mat', 'tvstand', 'closetdrawer', 'couch', 'clothes_hat', 'form', 'pillow', 'wall', 'nightstand', 'tape', 'television', 'after_shave', 'rag', 'diary', 'headset', 'sink', 'dishwasher', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'sponge', 'crayon', 'microwave', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'floor', 'bookshelf', 'button', 'stovefan', 'mat', 'stereo', 'mousepad', 'clothes_hat', 'bowl', 'nightstand', 'wall', 'bed', 'photoframe', 'light_bulb', 'food_food', 'television', 'kitchen_counter', 'dresser', 'door', 'sink', 'clothes_skirt', 'scrabble']\nTask:Breakfast\nresponse:\n{\n    \"thoughts\": I have breakfast in bed, watching tv.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[GRAB] <food_food>', '[WALK] <bedroom>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <bed>', '[LIE] <bed>', '[TURNTO] <television>', '[WATCH] <television>', '[EAT] <food_food>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'freezer', 'ground_coffee', 'bathroom', 'walllamp', 'detergent', 'food_butter', 'floor', 'ceiling', 'computer', 'food_bacon', 'bookmark', 'toaster', 'shoe-shine_kit', 'faucet', 'stereo', 'window', 'wall', 'coffe_maker', 'coffee_filter', 'water', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'instrument_guitar', 'table', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'dog', 'chair', 'bathroom', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'mat', 'ceilinglamp', 'shower', 'candle', 'wall', 'bed', 'clothes_gloves', 'food_food', 'water', 'food_rice', 'wooden_spoon', 'curtain', 'console', 'dishrack', 'pajamas', 'light', 'clothes_skirt', 'soap', 'dirt']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['fork', 'scissors', 'chair', 'drawing', 'wallshelf', 'floor', 'bookshelf', 'envelope', 'piano_bench', 'toaster', 'ceilinglamp', 'plate', 'shower', 'picture', 'oil', 'closetdrawer', 'microphone', 'food_cheese', 'newspaper', 'bed', 'food_food', 'wine', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'light']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.", "actions": ["[WALK] <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'freezer', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'microphone', 'pillow', 'wall', 'wall_clock', 'nightstand', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'home_office', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room and sit on the couch, ease into the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'bathroom', 'wallshelf', 'phone', 'bookshelf', 'button', 'mat', 'ceilinglamp', 'window', 'couch', 'pillow', 'dining_room', 'nightstand', 'photoframe', 'coffe_maker', 'ice', 'lighter', 'mouse', 'pajamas', 'cupboard', 'door', 'sink', 'spoon', 'clothes_shirt', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go into the living room and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'character', 'computer', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'window', 'deck_of_cards', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'food_kiwi', 'nightstand', 'wall', 'cup', 'coffe_maker', 'television', 'food_rice', 'curtain', 'dresser', 'napkin', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit on my couch and watch TV in my living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'dry_pasta', 'crayon', 'band-aids', 'detergent', 'drawing', 'clothes_underwear', 'table_cloth', 'stovefan', 'laptop', 'toaster', 'faucet', 'ceilinglamp', 'closetdrawer', 'wall', 'nightstand', 'oven', 'keys', 'food_food', 'tray', 'coffee_pot', 'pencil', 'headset', 'door', 'hairdryer', 'sink', 'table', 'dishwasher', 'clothes_shirt', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "You walk to the living room and find the cat. You look at the cat and touch it to pet it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'beer', 'tea', 'freezer', 'walllamp', 'nail_polish', 'bathroom', 'needle', 'cpuscreen', 'keyboard', 'mail', 'computer', 'stovefan', 'ceilinglamp', 'iron', 'clothes_pants', 'food_orange', 'pillow', 'dining_room', 'bench', 'bed', 'coffe_maker', 'television', 'doorjamb', 'pencil', 'maindoor', 'napkin', 'instrument_violin']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office, find where the phone is and lift the reciever.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'bathroom', 'knifeblock', 'wallshelf', 'bathroom_counter', 'character', 'bedroom', 'tea_bag', 'keyboard', 'food_cake', 'toaster', 'ceilinglamp', 'shower', 'faucet', 'couch', 'food_jam', 'clothes_pants', 'bowl', 'dining_room', 'nightstand', 'oven', 'ice', 'lighter', 'mouse', 'conditioner', 'door', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the living room, Find Phone stand. Find Phone. Pickup the receiver to attend the call.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'food_cake', 'stovefan', 'mat', 'toaster', 'shower', 'food_noodles', 'oil', 'nightstand', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>']\n}\n", "input": "\nScene_Objects:['beer', 'toilet', 'bathroom', 'detergent', 'needle', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'shampoo', 'ceiling', 'character', 'desk', 'bedroom', 'toaster', 'tvstand', 'window', 'clothes_socks', 'food_jam', 'pillow', 'towel', 'dining_room', 'cup', 'bed', 'food_food', 'coffe_maker', 'toilet_paper', 'television', 'tray', 'doorjamb', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the living room and then I walk to the light switch in there and turn the lights on", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'keyboard', 'mat', 'ceilinglamp', 'shower', 'oil', 'food_peanut_butter', 'food_orange', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'pencil', 'dresser', 'door', 'napkin', 'light', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": put away groceries in fridge. open the door, put on shelf, close the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['food_salt', 'beer', 'microwave', 'chair', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'mop_bucket', 'toaster', 'brush', 'plate', 'faucet', 'stereo', 'ceilinglamp', 'tvstand', 'dish_soap', 'wall', 'food_food', 'video_game_controller', 'toothbrush_holder', 'doorjamb', 'mouse', 'kitchen_counter', 'headset', 'cupboard', 'sink', 'light', 'table']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "i go to the kitchen, i find the faucet. I turn the faucet on, i grab the plate to wash , i then get the soap. I wash the dishes by hand", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'toaster', 'closetdrawer', 'couch', 'food_jam', 'pillow', 'wall', 'cup', 'photoframe', 'coffee_table', 'novel', 'television', 'doorjamb', 'shoes', 'comb', 'dresser', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'detergent', 'orchid', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'computer', 'tvstand', 'faucet', 'plate', 'window', 'oil', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'bed', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "turn on the tv", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'ceilinglamp', 'tvstand', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'clothes_gloves', 'novel', 'tray', 'doorjamb', 'kitchen_counter', 'door', 'table', 'light', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'bathroom', 'detergent', 'dough', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'bookshelf', 'toaster', 'mat', 'ceilinglamp', 'shower', 'laptop', 'dining_room', 'wall', 'cup', 'newspaper', 'food_food', 'toothbrush_holder', 'doorjamb', 'diary', 'kitchen_counter', 'door', 'sink', 'light', 'clothes_skirt', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch. Turn on lights.", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'food_bread', 'filing_cabinet', 'bathroom', 'ceiling', 'phone', 'keyboard', 'purse', 'bookshelf', 'laptop', 'ceilinglamp', 'faucet', 'nightstand', 'bench', 'photoframe', 'bed', 'clothes_gloves', 'food_food', 'coffe_maker', 'television', 'toothbrush_holder', 'curtain', 'shaving_cream', 'pencil', 'dresser', 'conditioner', 'door', 'spoon', 'bathroom_cabinet', 'table']\nTask:Shut front door\nresponse:\n{\n    \"thoughts\": Close the storm door. Lock the storm door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'chair', 'dog', 'freezer', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'pasta', 'light_bulb', 'clothes_gloves', 'television', 'curtain', 'doorjamb', 'conditioner', 'door', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and turn on the light. I walk to the couch then sit on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'freezer', 'cat', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'iron', 'balanceball', 'mousepad', 'dustpan', 'pillow', 'broom', 'wall', 'oven', 'coffee_table', 'novel', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'mouse', 'pencil', 'light', 'board_game']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I walk to the living room and pick up a book from the coffee table. I find my favorite chair.  I sit in the armchair and open the book.  I read.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[WALK] <coffee_table>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'food_butter', 'knifeblock', 'floor', 'ceiling', 'desk', 'tea_bag', 'bookshelf', 'mop_bucket', 'stovefan', 'ceilinglamp', 'faucet', 'stereo', 'window', 'closetdrawer', 'couch', 'pillow', 'bowl', 'wall', 'nightstand', 'oven', 'food_food', 'television', 'chessboard', 'sink', 'dishwasher', 'laser_pointer']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room. Walk to radio. Switch on the radio. Walk to sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'food_carrot', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'television', 'oven_mitts', 'wooden_spoon', 'doorjamb', 'mouse', 'pajamas', 'door', 'light', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Compose an email in the home office.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['scissors', 'microwave', 'dog', 'freezer', 'detergent', 'dough', 'knifeblock', 'bills', 'floor', 'tea_bag', 'thread', 'toaster', 'plate', 'faucet', 'dish_soap', 'wall', 'cutting_board', 'tape', 'food_cheese', 'bed', 'newspaper', 'coffe_maker', 'lighter', 'shaving_cream', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'light']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and see dishes in the sink. I go to the sink, fill the sink with soap and water, and wash the dishes. Then I rinse the dishes with clean water and put them in the drying rack.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[TURNTO] <plate>", "[LOOKAT] <plate>", "[WALK] <sink>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <sink>", "[WASH] <plate>", "[RINSE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'microwave', 'toilet', 'chair', 'walllamp', 'cpuscreen', 'hanger', 'phone', 'computer', 'stereo', 'faucet', 'window', 'couch', 'dining_room', 'wall_clock', 'bench', 'pasta', 'coffe_maker', 'food_food', 'doorjamb', 'mouse', 'pajamas', 'controller', 'sink', 'napkin', 'light', 'table', 'dishwasher']\nTask:Walk to room\nresponse:\n{\n    \"thoughts\": Leave the living room and enter the hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['sponge', 'water_glass', 'tea', 'microwave', 'chair', 'food_carrot', 'cpuscreen', 'kitchen_cabinet', 'floor', 'purse', 'computer', 'stovefan', 'faucet', 'food_noodles', 'wall', 'nightstand', 'oven', 'wall_clock', 'bed', 'water', 'after_shave', 'centerpiece', 'diary', 'pencil', 'kitchen_counter', 'headset', 'pajamas', 'table', 'dishwasher', 'board_game']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to kitchen. Open the the cabinet. Locate the water glass. Pour water into glass. Drag the bottle close to mouth. Drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[CLOSE] <kitchen_cabinet>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'towel', 'wall', 'wall_clock', 'oven', 'television', 'tray', 'rag', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <bathroom_cabinet>', '[FIND] <shower>', '[RINSE] <shower>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'toilet', 'walllamp', 'cat', 'orchid', 'floor', 'ceiling', 'remote_control', 'computer', 'food_bacon', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'mouthwash', 'television', 'coin', 'food_rice', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'table', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'toilet', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'ceiling', 'cleaning_solution', 'mop_bucket', 'table_cloth', 'faucet', 'shower', 'iron', 'clothes_socks', 'pillow', 'wall', 'food_food', 'mirror', 'wine', 'doorjamb', 'coffee_pot', 'paper_towel', 'cupboard', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'dishwasher', 'scrabble']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[FIND] <cleaning_solution>', '[RINSE] <cleaning_solution>', '[WIPE] <mirror>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'knifeblock', 'knife', 'hanger', 'floor', 'folder', 'desk', 'toaster', 'plate', 'faucet', 'tvstand', 'dish_soap', 'bowl', 'wall', 'cup', 'keys', 'coffe_maker', 'food_onion', 'kitchen_counter', 'dishrack', 'hairbrush', 'cupboard', 'sink', 'spoon', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the dishwasher. I open it. I find the sink and grab the dirty plates, cups, bowls, and utensils. I rinse off the dirty dishes under the faucet. I then place the dirty dishes into the dishwasher. I add soap to the dishwasher and close the lid. I turn it on and wait.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <sink>", "[FIND] <bowl>", "[GRAB] <bowl>", "[RINSE] <bowl>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[RINSE] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[RINSE] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[RINSE] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[RINSE] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[TURNTO] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'detergent', 'floor', 'ceiling', 'toaster', 'faucet', 'stereo', 'electric_shaver', 'window', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'cutting_board', 'coffe_maker', 'television', 'drinking_glass', 'after_shave', 'shaving_cream', 'pencil', 'kitchen_counter', 'comb', 'cupboard', 'sink', 'bathroom_cabinet', 'vase', 'clothes_shirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'computer', 'button', 'tvstand', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'oven', 'coffee_table', 'novel', 'television', 'curtain', 'doorjamb', 'lighter', 'dresser', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Go to the living room, find the remote control on the coffee table, and change the TV channel.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <coffee_table>", "[WALK] <coffee_table>", "[TURNTO] <coffee_table>", "[LOOKAT] <coffee_table>", "[FIND] <remote_control>", "[TOUCH] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[TOUCH] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'mat', 'shower', 'window', 'food_kiwi', 'wall', 'bench', 'newspaper', 'food_food', 'coffe_maker', 'food_onion', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'home_office']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'measuring_cup', 'freezer', 'bathroom', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'desk', 'piano_bench', 'stovefan', 'mat', 'faucet', 'stereo', 'couch', 'dining_room', 'photoframe', 'coffe_maker', 'toothbrush_holder', 'tray', 'doorjamb', 'console', 'mouse', 'dresser', 'conditioner', 'table', 'clothes_shirt', 'creditcard']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'sponge', 'scissors', 'chair', 'dog', 'drawing', 'bills', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'ceilinglamp', 'shoe-shine_kit', 'window', 'foundation', 'couch', 'wall', 'bench', 'basket_for_clothes', 'bed', 'novel', 'television', 'kitchen_counter', 'dresser', 'cupboard', 'bathroom_cabinet', 'light', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Grab book.  Read it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <dresser>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'crayon', 'microwave', 'toilet', 'chair', 'dough', 'drawing', 'knifeblock', 'clothes_underwear', 'floor', 'folder', 'bedroom', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'candle', 'clothes_hat', 'wall', 'food_food', 'food_onion', 'cards', 'mouse', 'kitchen_counter', 'cupboard', 'conditioner', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'dirt']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I walk into the kitchen to the sink. I put dishes in sink and wash them with my hands", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <sink>", "[PUTBACK] <plate> <sink>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'tablelamp', 'detergent', 'cat', 'knife', 'drawing', 'hanger', 'bathroom_counter', 'desk', 'food_cake', 'purse', 'envelope', 'piano_bench', 'stovefan', 'toaster', 'stereo', 'balanceball', 'floor_lamp', 'powersocket', 'love_seat', 'newspaper', 'bed', 'photoframe', 'video_game_controller', 'diary', 'maindoor', 'light', 'soap', 'board_game']\nTask:Read\nresponse:\n{\n    \"thoughts\": Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'food_ice_cream', 'microwave', 'walllamp', 'orchid', 'bills', 'knife', 'wallshelf', 'floor', 'ceiling', 'remote_control', 'computer', 'tvstand', 'faucet', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'light_bulb', 'coffe_maker', 'television', 'curtain', 'lighter', 'vacuum_cleaner', 'dresser', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I grab the remote.  I find the red power circular button. I point the remote at the television.   I push button down and TV screen lights up. I sit and watch.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'ground_coffee', 'drawing', 'orchid', 'cpuscreen', 'phone', 'bedroom', 'stamp', 'desk', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'nightstand', 'wall', 'oven', 'wall_clock', 'photoframe', 'coffe_maker', 'mouse', 'dresser', 'cupboard', 'bathtub', 'sink', 'home_office', 'table', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": You walk to the living room and find the cat. You look at the cat and touch it to pet it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'bathroom', 'orchid', 'bills', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'bowl', 'nightstand', 'wall', 'oven', 'bed', 'novel', 'food_food', 'television', 'oven_mitts', 'curtain', 'dresser', 'controller', 'table', 'dishwasher', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room ,find television , switch on remote control", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[SWITCHON] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'trashcan', 'toothbrush', 'bathroom_counter', 'shampoo', 'floor', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'toaster', 'ceilinglamp', 'iron', 'shower', 'stereo', 'pillow', 'wall', 'wall_clock', 'photoframe', 'food_food', 'water', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[FIND] <toothbrush>', '[FIND] <water>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'cd_player', 'toilet', 'blow_dryer', 'floor', 'desk', 'tea_bag', 'remote_control', 'button', 'toaster', 'plate', 'tvstand', 'couch', 'nightstand', 'cup', 'photoframe', 'keys', 'novel', 'check', 'diary', 'vacuum_cleaner', 'cd', 'comb', 'hair', 'clothes_scarf', 'home_office', 'light', 'table', 'creditcard']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[PLUGIN] <cd_player>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'centerpiece', 'mouse', 'cupboard', 'sink', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Go find a computer, turn it on, click google, and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'knifeblock', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'faucet', 'couch', 'dustpan', 'pillow', 'towel', 'wall', 'nightstand', 'oven', 'box', 'bed', 'keys', 'check', 'toilet_paper', 'television', 'kitchen_counter', 'dresser', 'instrument_guitar', 'soap', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to living room.  Find sofa.  Sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'microwave', 'freezer', 'ground_coffee', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'piano_bench', 'shoe-shine_kit', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'tape', 'food_food', 'doorjamb', 'headset', 'cupboard', 'door', 'home_office', 'table', 'light', 'bathroom_cabinet', 'board_game']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I get home from work and walk into the living room. I see the couch and go to sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'bathroom', 'food_carrot', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'folder', 'desk', 'bookshelf', 'computer', 'mat', 'window', 'clothes_socks', 'wall', 'powersocket', 'toy', 'bed', 'food_food', 'television', 'doorjamb', 'hairbrush', 'cupboard', 'door', 'hairdryer', 'bathroom_cabinet', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find light switch.  Switch on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'toilet', 'chair', 'bathroom', 'food_carrot', 'cpuscreen', 'shampoo', 'phone', 'ceiling', 'bedroom', 'desk', 'computer', 'toaster', 'shower', 'foundation', 'dustpan', 'photoframe', 'coffe_maker', 'food_rice', 'tray', 'curtain', 'doorjamb', 'console', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher', 'creditcard']\nTask:Put umbrella away\nresponse:\n{\n    \"thoughts\": Place the umbrella in the umbrella bucket point first.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mop_bucket>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'band-aids', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'oven', 'newspaper', 'television', 'curtain', 'headset', 'dresser', 'clothes_dress', 'spoon', 'table', 'dishwasher', 'clothes_skirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "change tv channel with the remote, then watch the tv", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TOUCH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'walllamp', 'nail_polish', 'detergent', 'dough', 'drawing', 'food_egg', 'tea_bag', 'bookshelf', 'remote_control', 'alcohol', 'button', 'mop_bucket', 'table_cloth', 'stovefan', 'tvstand', 'shoe-shine_kit', 'shower', 'stereo', 'deck_of_cards', 'mousepad', 'powersocket', 'check', 'rag', 'towel_rack', 'pencil', 'conditioner', 'door', 'bathroom_cabinet', 'soap']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['glue', 'dry_pasta', 'food_ice_cream', 'chair', 'walllamp', 'nail_polish', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'photoframe', 'toilet_paper', 'television', 'wine', 'doorjamb', 'dresser', 'clothes_shirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "We recently got a 43 inch TV and have enjoyed watching some of our favorite movies and TV shows. We use a remote to do so.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'blender', 'tea', 'toilet', 'chair', 'freezer', 'bathroom', 'woman', 'knifeblock', 'drawing', 'wallshelf', 'floor', 'phone', 'character', 'desk', 'man', 'button', 'deck_of_cards', 'mousepad', 'clothes_pants', 'child', 'coffe_maker', 'tray', 'comb', 'dresser', 'cupboard', 'hairbrush', 'soap', 'tooth_paste']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Greet guests when they knock on the door and invite them to enter my home,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['beer', 'microwave', 'toilet', 'walllamp', 'ground_coffee', 'detergent', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'computer', 'alcohol', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'wall', 'photoframe', 'keys', 'television', 'food_rice', 'tray', 'doorjamb', 'shaving_cream', 'home_office']\nTask:Text friends while sitting on couch\nresponse:\n", "output": {"thoughts": "Text while sitting", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[TOUCH] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ceilingfan', 'toothbrush', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'longboard', 'mop_bucket', 'mat', 'tvstand', 'faucet', 'couch', 'clothes_pants', 'towel', 'wall', 'basket_for_clothes', 'photoframe', 'light_bulb', 'wooden_spoon', 'after_shave', 'coffee', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'laser_pointer', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Turn faucet clockwise to turn on water.  Place hands under stream of water, pick up soap, rub hands with soap.  Put soap back down.  Rub hands together with soap on them.  Rinse hands under water from faucet.  Turn faucet counterclockwise to turn water off.  Dry hands by rubbing them on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[WASH] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['chair', 'band-aids', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'homework', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'keys', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'light', 'soap']\nTask:Change TV channels\nresponse:\n", "output": {"thoughts": "Take remote. Change to desired channel by you.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[TOUCH] <remote_control>", "[TOUCH] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'cat', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'oil', 'closetdrawer', 'food_orange', 'form', 'pillow', 'wall', 'nightstand', 'powersocket', 'tape', 'photoframe', 'doorjamb', 'console', 'kitchen_counter', 'door', 'light', 'dishwasher', 'clothes_skirt', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": approach the light switch, raise hand to the switch, flip the switch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['spectacles', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'rag', 'doorjamb', 'cards', 'dresser', 'sink', 'dishwasher']\nTask:Watch  TV\nresponse:\n", "output": {"thoughts": "Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'knifeblock', 'knife', 'hanger', 'floor', 'folder', 'desk', 'toaster', 'plate', 'faucet', 'tvstand', 'dish_soap', 'bowl', 'wall', 'cup', 'keys', 'coffe_maker', 'food_onion', 'kitchen_counter', 'dishrack', 'hairbrush', 'cupboard', 'sink', 'spoon', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the dishwasher. I open it. I find the sink and grab the dirty plates, cups, bowls, and utensils. I rinse off the dirty dishes under the faucet. I then place the dirty dishes into the dishwasher. I add soap to the dishwasher and close the lid. I turn it on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <sink>', '[FIND] <bowl>', '[GRAB] <bowl>', '[RINSE] <bowl>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[RINSE] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[RINSE] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[RINSE] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[RINSE] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[TURNTO] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sponge', 'fork', 'filing_cabinet', 'tablelamp', 'chair', 'band-aids', 'floor', 'ceiling', 'stamp', 'alcohol', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'towel', 'wall', 'toy', 'bed', 'keys', 'photoframe', 'coffe_maker', 'television', 'coin', 'kitchen_counter', 'dresser', 'cupboard', 'food_steak', 'sink', 'table']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Run warm water in the sink. Pour dish washing liquid on the dishes. Pick up dishes and wash them individually with a sponge. Rinse suds off under running water. Dry dishes with towel.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[WASH] <plate>", "[RINSE] <plate>", "[RINSE] <sponge>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <sponge>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <plate>", "[PUTOBJBACK] <towel>", "[WALK] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[PUTIN] <plate> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'trashcan', 'orchid', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'window', 'food_noodles', 'couch', 'pillow', 'wall', 'photoframe', 'toothbrush_holder', 'curtain', 'doorjamb', 'chessboard', 'door', 'bathtub', 'bag', 'home_office', 'light', 'table', 'dishwasher', 'instrument_violin']\nTask:Admire art\nresponse:\n{\n    \"thoughts\": Looking at painting, art,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'ground_coffee', 'food_butter', 'food_carrot', 'bills', 'floor', 'ceiling', 'bedroom', 'computer', 'table_cloth', 'tvstand', 'shower', 'faucet', 'wall', 'oven', 'newspaper', 'photoframe', 'bed', 'coffee_filter', 'curtain', 'kitchen_counter', 'cupboard', 'controller', 'dishwasher', 'clothes_shirt']\nTask:Make popcorn\nresponse:\n", "output": {"thoughts": "Lay popcorn bag flat inside microwave and press popcorn button to start.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <microwave>", "[TURNTO] <microwave>", "[OPEN] <microwave>", "[PUTIN] <food_food> <microwave>", "[CLOSE] <microwave>", "[SWITCHON] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'microwave', 'chair', 'freezer', 'dog', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'phone', 'dvd_player', 'character', 'bookshelf', 'computer', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'bench', 'photoframe', 'food_food', 'tray', 'curtain', 'kitchen_counter', 'door', 'sink', 'table']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['tea', 'microwave', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'button', 'food_bacon', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'coffee_table', 'novel', 'mouthwash', 'television', 'doorjamb', 'diary', 'clothes_scarf', 'sink']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit down in recliner.  Pick up novel off of coffee table.  Open novel to last read page.  Begin reading", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'bathroom', 'detergent', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'video_game_console', 'keyboard', 'stovefan', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall_clock', 'oven', 'bed', 'coffe_maker', 'video_game_controller', 'television', 'doorjamb', 'cd', 'cupboard', 'door', 'bathtub', 'bathroom_cabinet', 'light', 'home_office', 'table']\nTask:Playing video game\nresponse:\n{\n    \"thoughts\": Playing  video games  on various video game  systems such as wii u  and  xbox,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <video_game_console>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[PUTOBJBACK] <video_game_controller>', '[SWITCHOFF] <video_game_console>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'chair', 'blow_dryer', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'toaster', 'laptop', 'deck_of_cards', 'faucet', 'food_noodles', 'closetdrawer', 'food_orange', 'wall', 'nightstand', 'clothes_gloves', 'coffe_maker', 'toilet_paper', 'food_food', 'curtain', 'towel_rack', 'console', 'kitchen_counter', 'shoes', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'chair', 'walllamp', 'knifeblock', 'floor', 'phone', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'purse', 'toaster', 'mat', 'stereo', 'faucet', 'mousepad', 'pillow', 'mouthwash', 'curtain', 'cards', 'shoes', 'cd', 'dishrack', 'dresser', 'door', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Read a book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'toy', 'photoframe', 'bed', 'television', 'doorjamb', 'towel_rack', 'dresser', 'dishwasher', 'shoe_rack']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sit down on couch in front of TV.  Use remote to turn on the TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'jelly', 'chair', 'walllamp', 'detergent', 'trashcan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'keyboard', 'stovefan', 'mat', 'ceilinglamp', 'food_orange', 'wall', 'bench', 'oven', 'food_food', 'oven_mitts', 'doorjamb', 'dresser', 'cupboard', 'home_office', 'table', 'bathroom_cabinet']\nTask:Play games\nresponse:\n{\n    \"thoughts\": I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <board_game>', '[FIND] <board_game>', '[GRAB] <board_game>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <board_game> <table>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'chair', 'detergent', 'toothbrush', 'floor', 'ceiling', 'food_cake', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'ceilinglamp', 'dish_soap', 'closetdrawer', 'pillow', 'bowl', 'dining_room', 'wall', 'powersocket', 'cup', 'food_cheese', 'clothes_gloves', 'coffe_maker', 'after_shave', 'kitchen_counter', 'dishrack', 'cupboard', 'sink']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Turn water on and run over dishes while scrubbing with a soapy sponge.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'microwave', 'chair', 'band-aids', 'woman', 'needle', 'longboard', 'man', 'keyboard', 'envelope', 'toaster', 'mat', 'ceilinglamp', 'shower', 'faucet', 'stereo', 'nightstand', 'box', 'light_bulb', 'coffe_maker', 'curtain', 'towel_rack', 'shoes', 'dresser', 'hairbrush', 'door', 'vase', 'instrument_violin', 'creditcard']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'beer', 'microwave', 'chair', 'freezer', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'stovefan', 'shower', 'wall', 'powersocket', 'oven', 'coffe_maker', 'food_food', 'tray', 'rag', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'table', 'dishwasher', 'light']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Take out a rag. Turn on faucet and get water on the rag. Clean counter off.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <rag> <sink>", "[SWITCHOFF] <faucet>", "[GRAB] <rag>", "[SQUEEZE] <rag>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'table_cloth', 'ceilinglamp', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'shelf', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office', 'light', 'table', 'laser_pointer']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Find the computer.  Turn on the computer.  Open the e-mail program.  Create a new message.  Write the message.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'mop', 'freezer', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'table_cloth', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'keys', 'newspaper', 'food_rice', 'doorjamb', 'dresser', 'door', 'bathroom_cabinet', 'light', 'scrabble']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'freezer', 'bathroom', 'detergent', 'food_butter', 'longboard', 'bookshelf', 'stovefan', 'shower', 'pillow', 'towel', 'dining_room', 'nightstand', 'powersocket', 'toy', 'basket_for_clothes', 'newspaper', 'light_bulb', 'photoframe', 'food_food', 'coin', 'doorjamb', 'headset', 'shoes', 'hairbrush', 'cupboard', 'napkin', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into my room and turn the radio on so that I can listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'toilet', 'chair', 'freezer', 'walllamp', 'toothbrush', 'phone', 'character', 'bedroom', 'bookshelf', 'purse', 'food_cake', 'stovefan', 'stereo', 'electric_shaver', 'homework', 'couch', 'dining_room', 'nightstand', 'toy', 'light_bulb', 'video_game_controller', 'wooden_spoon', 'tray', 'kitchen_counter', 'sink', 'clothes_scarf', 'home_office', 'bag', 'light']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'keyboard', 'purse', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'food_noodles', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": FIRST YOU NEED TO SWITVH ON YOU COMPUTER AND CHECK THE INTERNET CONNECTION.FOR SENDING EMAIL EMAIL ID FOR REQUIRED FOR BOTH SEND AND RECEIVE .SO CLEALY MENTION THE ID ON THE TO PLACE THEN ONLY IT WILL GO CORRECTLY,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Draft home\nresponse:\n", "output": {"thoughts": "Using mathematics and CAD software design a home", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[POINTAT] <computer>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'food_carrot', 'knife', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'stamp', 'purse', 'computer', 'mat', 'shoe-shine_kit', 'shower', 'foundation', 'mousepad', 'closetdrawer', 'food_jam', 'powersocket', 'coffe_maker', 'television', 'tray', 'curtain', 'diary', 'cards', 'pencil', 'table']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to the radio.  I switch on the radio and listen,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'remote_control', 'computer', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'clothes_socks', 'closetdrawer', 'couch', 'pillow', 'towel', 'wall', 'powersocket', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch and use the remote control to turn on the TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'ground_coffee', 'blow_dryer', 'orchid', 'wallshelf', 'bathroom_counter', 'character', 'desk', 'bookshelf', 'alcohol', 'computer', 'ceilinglamp', 'laptop', 'faucet', 'pillow', 'wall_clock', 'powersocket', 'oven', 'cutting_board', 'cup', 'television', 'oven_mitts', 'after_shave', 'doorjamb', 'diary', 'towel_rack', 'cards', 'maindoor']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find cat. Walk to cat. Pull cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[PULL] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'blow_dryer', 'walllamp', 'detergent', 'floor', 'ceiling', 'bedroom', 'keyboard', 'purse', 'tvstand', 'deck_of_cards', 'clothes_pants', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'keys', 'tray', 'kettle', 'shoes', 'hair', 'sink', 'home_office', 'table', 'dishwasher', 'soap', 'board_game', 'dirt']\nTask:Hang up car keys\nresponse:\n", "output": {"thoughts": "Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <purse>", "[FIND] <purse>", "[OPEN] <purse>", "[FIND] <keys>", "[GRAB] <keys>", "[CLOSE] <purse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'faucet', 'window', 'oil', 'couch', 'wall', 'bench', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'after_shave', 'towel_rack', 'kitchen_counter', 'bathtub', 'sink', 'light']\nTask:Open bathroom window\nresponse:\n{\n    \"thoughts\": Make sure bathroom window is unlocked. Lift up bathroom window so it is open.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <window>', '[FIND] <window>', '[OPEN] <window>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'floor_lamp', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'door', 'light', 'laser_pointer']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Take a book off the bookshelf. Sit down in a chair at the desk. Turn on desk lamp. Read the book.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'nail_polish', 'bathroom', 'cat', 'dough', 'toothbrush', 'desk', 'button', 'stovefan', 'shower', 'faucet', 'window', 'pillow', 'nightstand', 'wall_clock', 'oven', 'cutting_board', 'novel', 'check', 'television', 'toothbrush_holder', 'lighter', 'kettle', 'chessboard', 'dresser', 'sink', 'home_office', 'clothes_skirt']\nTask:Keep cats out of room\nresponse:\n{\n    \"thoughts\": Stand in the doorway and create a loud noise so the cats will stay out.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'picture', 'mousepad', 'wall', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'cards', 'mouse', 'door', 'light', 'printing_paper']\nTask:Add paper to printer\nresponse:\n", "output": {"thoughts": "Take a stack of paper, about 25 pieces. Place carefully in the paper tray.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <printing_paper>", "[GRAB] <printing_paper>", "[WALK] <fax_machine>", "[PUTBACK] <printing_paper> <fax_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_cake', 'computer', 'food_bacon', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'deck_of_cards', 'mousepad', 'couch', 'wall', 'nightstand', 'newspaper', 'television', 'mouse', 'door', 'sink', 'home_office']\nTask:Read newspaper\nresponse:\n{\n    \"thoughts\": Sit in my chair at my desk. Open up newspaper. Read about current events.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[OPEN] <newspaper>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to the computer, sit in the chair, pull up computer screen get the mouse and keyboard load work software", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <mousepad>", "[PUTBACK] <mouse> <mousepad>", "[TURNTO] <computer>", "[POINTAT] <computer>", "[PULL] <mouse>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <mouse>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'bathroom_cabinet']\nTask:Use computer\nresponse:\n{\n    \"thoughts\": Browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Open computer. Log in to email provider. Click new email tab. Write email. Click send.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <mail>", "[OPEN] <mail>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'bathroom', 'food_butter', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mop_bucket', 'toaster', 'shower', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'toy', 'bed', 'razor', 'television', 'after_shave', 'doorjamb', 'dresser', 'conditioner', 'door', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into the bedroom and walk to the light switch. Flip the switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'food_carrot', 'drawing', 'orchid', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'computer', 'mat', 'plate', 'ceilinglamp', 'stereo', 'couch', 'candle', 'food_orange', 'pillow', 'wall', 'oven', 'food_vegetable', 'novel', 'food_food', 'razor', 'television', 'console', 'kitchen_counter', 'cupboard', 'controller', 'table']\nTask:Fix snack\nresponse:\n", "output": {"thoughts": "Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <kitchen_cabinet>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTBACK] <food_carrot> <plate>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[PUTBACK] <food_vegetable> <plate>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'tablelamp', 'chair', 'freezer', 'detergent', 'food_carrot', 'knife', 'cpuscreen', 'kitchen_cabinet', 'character', 'bedroom', 'mail', 'stovefan', 'laptop', 'form', 'pillow', 'nightstand', 'bench', 'oven', 'cup', 'keys', 'toothbrush_holder', 'water', 'food_rice', 'curtain', 'controller', 'sink', 'bathtub', 'dishwasher', 'dirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Get a glass of water to drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'chair', 'walllamp', 'cat', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'bowl', 'wall', 'powersocket', 'box', 'photoframe', 'television', 'doorjamb', 'dresser', 'conditioner', 'door', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "We sit all around the television and watch a program.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[PUTBACK] <remote_control> <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'crayon', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'door', 'clothes_scarf', 'light']\nTask:Re arrange office\nresponse:\n{\n    \"thoughts\": Walk in.Turn on light.Move furniture. Change the room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[PUSH] <desk>', '[WALK] <chair>', '[PUSH] <chair>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Complete surveys on amazon turk\nresponse:\n", "output": {"thoughts": "Sit down at computer, Turn computer on. Go online to Amazon mechanical turk, Browse through surveys, Select a survey.", "actions": ["[WALK] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[POINTAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'cutting_board', 'cup', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'controller', 'door', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[PUTOBJBACK] <remote_control>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'band-aids', 'freezer', 'food_carrot', 'toothbrush', 'orchid', 'floor', 'mail', 'button', 'faucet', 'foundation', 'oil', 'pillow', 'towel', 'food_food', 'toothbrush_holder', 'doorjamb', 'lighter', 'kettle', 'headset', 'cd', 'dresser', 'hairbrush', 'controller', 'napkin', 'light', 'dishwasher', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the toothbrush holder, take toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now brush teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[TURNTO] <toothbrush>', '[LOOKAT] <toothbrush>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'television', 'tray', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch youtube\nresponse:\n", "output": {"thoughts": "Go to youtube.Com watch videos", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'toaster', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'water', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['sponge', 'crayon', 'chair', 'walllamp', 'hanger', 'floor', 'ceiling', 'phone', 'tea_bag', 'clothes_jacket', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'dish_soap', 'mousepad', 'dustpan', 'clothes_hat', 'wall', 'cup', 'coffe_maker', 'coffee_filter', 'curtain', 'coffee_pot', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'napkin']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sink>", "[PUTOBJBACK] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <cup>", "[GRAB] <cup>", "[SCRUB] <cup>", "[PUTOBJBACK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[SCRUB] <cup>", "[PUTOBJBACK] <cup>", "[PUTBACK] <sponge> <sink>", "[SWITCHON] <faucet>", "[RINSE] <cup>", "[GRAB] <cup>", "[FIND] <dishrack>", "[PUTBACK] <cup> <dishrack>", "[RINSE] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishrack>", "[RINSE] <sponge>", "[SWITCHOFF] <faucet>", "[GRAB] <sponge>", "[PUTOBJBACK] <sponge>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'freezer', 'walllamp', 'food_butter', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'alcohol', 'window', 'food_jam', 'dining_room', 'wall', 'bench', 'powersocket', 'newspaper', 'food_food', 'doorjamb', 'kettle', 'towel_rack', 'dresser', 'cupboard', 'door', 'table', 'light']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'juice', 'microwave', 'jelly', 'bathroom', 'cat', 'dough', 'drawing', 'hanger', 'phone', 'character', 'bookshelf', 'food_bacon', 'ceilinglamp', 'tvstand', 'shower', 'clothes_hat', 'bowl', 'oven', 'toy', 'bed', 'photoframe', 'razor', 'curtain', 'vacuum_cleaner', 'cupboard', 'bathroom_cabinet', 'teeth', 'electrical_outlet']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug vacuum into the wall. Turn on the vacuum. Move vacuum around the floor to clean it. Unplug vacuum.", "actions": ["[WALK] <home_office>", "[WALK] <vacuum_cleaner>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[FIND] <electrical_outlet>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[PUTOBJBACK] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'shower', 'window', 'homework', 'mousepad', 'pillow', 'powersocket', 'photoframe', 'coffe_maker', 'doorjamb', 'kettle', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'clothes_dress', 'door', 'bag', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to bedroom and grab dirty pants and take to the washing machine in entrance hall. Put clothes in washing machine and then put in laundry detergent.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_pants>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'food_noodles', 'closetdrawer', 'couch', 'candle', 'pillow', 'dining_room', 'wall', 'food_cheese', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'headset', 'dresser', 'creditcard']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch favorite shows, netflix, hulu", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTBACK] <remote_control> <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tea', 'chair', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'plate', 'mousepad', 'couch', 'clothes_hat', 'wall', 'nightstand', 'cup', 'food_food', 'check', 'coffe_maker', 'after_shave', 'rag', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'hairdryer', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": go to living room,find the text book on the self ,sat on the chair ,open and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['scissors', 'microwave', 'freezer', 'character', 'bedroom', 'computer', 'stovefan', 'mat', 'plate', 'homework', 'closetdrawer', 'clothes_hat', 'dining_room', 'powersocket', 'basket_for_clothes', 'cup', 'bed', 'keys', 'coffe_maker', 'check', 'television', 'towel_rack', 'dishrack', 'dresser', 'napkin', 'home_office', 'light', 'table', 'instrument_violin', 'shoe_rack']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Get a broom or duster with long handle. Dust blades of ceiling fan.", "actions": ["[WALK] <bedroom>", "[WALK] <duster>", "[FIND] <duster>", "[GRAB] <duster>", "[PUTOBJBACK] <duster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blow_dryer', 'trashcan', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookmark', 'toaster', 'stovefan', 'tvstand', 'food_jam', 'pillow', 'dining_room', 'bowl', 'wall', 'powersocket', 'clothes_gloves', 'novel', 'coffe_maker', 'coffee_filter', 'tray', 'after_shave', 'curtain', 'cd', 'chessboard', 'cupboard', 'sink', 'dishwasher', 'scrabble', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[WALK] <coffee>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door']\nTask:Watch youtube\nresponse:\n", "output": {"thoughts": "Go to youtube.Com watch videos", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'drawing', 'orchid', 'knifeblock', 'bills', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'shower', 'stereo', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'television', 'food_rice', 'tray', 'curtain', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'home_office', 'table', 'light', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the sofa. Sit down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'chair', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'stereo', 'floor_lamp', 'mousepad', 'wall', 'nightstand', 'novel', 'video_game_controller', 'television', 'mouse', 'door', 'napkin']\nTask:Study\nresponse:\n", "output": {"thoughts": "Study norwegian", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'character', 'bedroom', 'keyboard', 'piano_bench', 'bookmark', 'laptop', 'electric_shaver', 'food_kiwi', 'wall', 'powersocket', 'toothbrush_holder', 'curtain', 'rag', 'doorjamb', 'shaving_cream', 'cards', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'light', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the dining room and turn on the lights,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Surf web for money legitimate making opportunities\nresponse:\n", "output": {"thoughts": "Sit down in front of computer. Turn computer on. Go online. Surf the web.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'tablelamp', 'blender', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'mat', 'balanceball', 'clothes_socks', 'dustpan', 'wall', 'powersocket', 'food_cheese', 'after_shave', 'doorjamb', 'towel_rack', 'dresser', 'door', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Hit light switch,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'toilet', 'dog', 'freezer', 'bathroom', 'orchid', 'bathroom_counter', 'phone', 'character', 'tea_bag', 'bookshelf', 'keyboard', 'stovefan', 'mat', 'plate', 'mousepad', 'oil', 'wall', 'newspaper', 'novel', 'food_food', 'television', 'rag', 'lighter', 'pencil', 'shoes', 'sink', 'home_office']\nTask:Open window\nresponse:\n", "output": {"thoughts": "Unlock the latch at the top. Push the bottom half of the window upwards.", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'computer', 'ceilinglamp', 'faucet', 'teeth', 'towel', 'wall', 'nightstand', 'powersocket', 'basket_for_clothes', 'photoframe', 'light_bulb', 'toilet_paper', 'television', 'food_donut', 'tray', 'cards', 'towel_rack', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Hand washing\nresponse:\n{\n    \"thoughts\": Turn on the faucet. Soap up your hands. Then rinse under the running water of the faucet. Turn off faucet and dry hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[WASH] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I have my desktop computer set up on my desk and complete the paperwork I brought home with me from work.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>", "[FIND] <desk>", "[OPEN] <desk>", "[PUTBACK] <check> <desk>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'crayon', 'dog', 'toothbrush', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'faucet', 'shower', 'balanceball', 'mousepad', 'wall', 'pasta', 'powersocket', 'oven', 'nightstand', 'photoframe', 'video_game_controller', 'coffee_filter', 'wooden_spoon', 'console', 'shoes', 'sink', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": Put toothpaste on toothbrush, brush teeth, rinse mouth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>', '[PUTOBJBACK] <tooth_paste>', '[WALK] <sink>', '[FIND] <teeth>', '[SCRUB] <teeth>', '[SCRUB] <teeth>', '[SCRUB] <teeth>', '[PUTBACK] <toothbrush> <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[RINSE] <teeth>', '[SWITCHOFF] <faucet>', '[GRAB] <toothbrush>', '[PUTOBJBACK] <toothbrush>']\n}\n", "input": "\nScene_Objects:['blender', 'tea', 'microwave', 'orchid', 'bathroom_counter', 'character', 'desk', 'mat', 'shower', 'closetdrawer', 'couch', 'clothes_pants', 'food_orange', 'tape', 'cup', 'basket_for_clothes', 'bed', 'food_food', 'coin', 'wooden_spoon', 'towel_rack', 'shoes', 'dishrack', 'hairbrush', 'controller', 'door', 'sink', 'table', 'board_game', 'shoe_rack']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'ground_coffee', 'bathroom', 'trashcan', 'food_butter', 'knifeblock', 'needle', 'floor', 'ceiling', 'character', 'clothes_jacket', 'toaster', 'ceilinglamp', 'balanceball', 'mousepad', 'broom', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'rag', 'coffee_pot', 'laundry_detergent', 'maindoor', 'napkin', 'table', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'food_cake', 'envelope', 'computer', 'mop_bucket', 'laptop', 'mat', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'television', 'doorjamb', 'mouse', 'controller', 'door']\nTask:Surf internet\nresponse:\n", "output": {"thoughts": "SIT ON COUCH OR CHAIR AND SURF SHOPPING AN SOCIAL MEDIA SITES", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <couch>", "[SIT] <couch>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'beer', 'tea', 'freezer', 'shampoo', 'character', 'desk', 'piano_bench', 'ceilinglamp', 'iron', 'picture', 'mousepad', 'couch', 'bowl', 'nightstand', 'wall_clock', 'toy', 'bed', 'coffe_maker', 'television', 'coin', 'after_shave', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'shoes', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": take remote control, push \"change channel\" button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light', 'creditcard']\nTask:Shop\nresponse:\n", "output": {"thoughts": "find computer, find shopping website, pick items, get credit card, make a purchase", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <creditcard>", "[GRAB] <creditcard>", "[TURNTO] <creditcard>", "[LOOKAT] <creditcard>", "[FIND] <desk>", "[OPEN] <desk>", "[PUTBACK] <creditcard> <desk>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'toilet', 'chair', 'bathroom', 'detergent', 'dough', 'knife', 'drawing', 'floor', 'ceiling', 'tea_bag', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'mousepad', 'clothes_socks', 'pillow', 'wall', 'cup', 'doorjamb', 'console', 'shoes', 'dresser', 'door', 'hairdryer', 'napkin', 'light', 'soap', 'laser_pointer']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk over to light switch. Turn on lights.,\n    \"actions\": ['[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'nail_polish', 'cat', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'tray', 'doorjamb', 'dresser', 'cupboard', 'controller', 'door', 'sink', 'bathroom_cabinet', 'tooth_paste', 'creditcard']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on the couch and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'ground_coffee', 'walllamp', 'detergent', 'knifeblock', 'longboard', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'faucet', 'candle', 'wall', 'box', 'coffe_maker', 'food_food', 'television', 'coffee_filter', 'water', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'hair', 'maindoor', 'sink', 'napkin', 'dishwasher', 'tooth_paste']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink']\nTask:Use computer\nresponse:\n", "output": {"thoughts": "Browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'mop', 'ground_coffee', 'ceilingfan', 'trashcan', 'toothbrush', 'drawing', 'needle', 'bookshelf', 'keyboard', 'laptop', 'balanceball', 'stereo', 'mousepad', 'wall_clock', 'bench', 'coffe_maker', 'food_food', 'toothbrush_holder', 'television', 'curtain', 'doorjamb', 'kettle', 'controller', 'napkin', 'home_office', 'light', 'vase', 'soap', 'tooth_paste']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the toothbrush holder, take toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now brush teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[TURNTO] <toothbrush>', '[LOOKAT] <toothbrush>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'freezer', 'trashcan', 'drawing', 'bathroom_counter', 'ceiling', 'phone', 'thread', 'computer', 'stovefan', 'faucet', 'shower', 'food_jam', 'dining_room', 'wall', 'powersocket', 'tape', 'coffe_maker', 'food_food', 'mirror', 'doorjamb', 'mouse', 'paper_towel', 'dresser', 'cupboard', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'home_office', 'dirt']\nTask:Clean  mirror\nresponse:\n", "output": {"thoughts": "Spray Windex or glass and mirror cleaner onto mirror. Use paper towel to wipe off cleaner until mirror is dry and has no streaks.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[CLOSE] <bathroom_cabinet>", "[WALK] <mirror>", "[WIPE] <mirror>", "[FIND] <trashcan>", "[PUTBACK] <paper_towel> <trashcan>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": Find home office. Walk to home office. Find computer. Find chair at desk. Sit in chair. Switch on computer. Work on computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'bookmark', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'mousepad', 'dining_room', 'wall', 'nightstand', 'oven', 'cup', 'bed', 'television', 'coffee_filter', 'mouse', 'door']\nTask:Answer emails\nresponse:\n", "output": {"thoughts": "Open email application.  Open new emails.  Respond accordingly", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <mail>", "[GRAB] <mail>", "[READ] <mail>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'bathroom', 'detergent', 'cat', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'mousepad', 'form', 'wall', 'powersocket', 'photoframe', 'food_food', 'check', 'coffee_filter', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'controller', 'door', 'light', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'candle', 'pillow', 'wall', 'photoframe', 'bed', 'television', 'food_rice', 'doorjamb', 'centerpiece', 'dresser', 'napkin', 'light', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on the couch and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'bathroom', 'detergent', 'orchid', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'purse', 'computer', 'table_cloth', 'bookmark', 'stovefan', 'stereo', 'oil', 'closetdrawer', 'coffe_maker', 'television', 'food_donut', 'food_rice', 'tray', 'centerpiece', 'doorjamb', 'pencil', 'maindoor', 'sink', 'home_office', 'table', 'clothes_skirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to living room. Find book. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_bread', 'juice', 'food_salt', 'toilet', 'walllamp', 'detergent', 'dough', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'character', 'tea_bag', 'toaster', 'mat', 'broom', 'powersocket', 'love_seat', 'newspaper', 'photoframe', 'keys', 'food_food', 'coffe_maker', 'toilet_paper', 'ice', 'tray', 'lighter', 'kettle', 'home_office', 'dirt']\nTask:Read\nresponse:\n", "output": {"thoughts": "Reading a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'purse', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'electric_shaver', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office. Sit on chair. Switch on computer. type in keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Compute\nresponse:\n", "output": {"thoughts": "Use computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'scissors', 'drawing', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'shower', 'electric_shaver', 'food_kiwi', 'wall', 'bench', 'food_cheese', 'photoframe', 'newspaper', 'bed', 'food_food', 'standingmirror', 'television', 'curtain', 'doorjamb', 'shaving_cream', 'door', 'soap', 'clothes_shirt', 'scrabble', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control,\n    \"actions\": ['[WALK] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'thread', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'wall_clock', 'basket_for_clothes', 'photoframe', 'television', 'food_onion', 'doorjamb', 'dresser', 'door', 'clothes_scarf']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch your favorite TV shows.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'laptop', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dishrack', 'door', 'light', 'shoe_rack']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office, walk to desk, sit in chair, turn on laptop, grab mouse, open a web page to look at it, move mouse, keep looking at laptop screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <mouse>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'bathroom_cabinet']\nTask:Use computer\nresponse:\n", "output": {"thoughts": "Browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TURNTO] <computer>", "[POINTAT] <computer>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'freezer', 'bathroom', 'ceilingfan', 'drawing', 'bills', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'toy', 'tape', 'bed', 'television', 'curtain', 'comb', 'hair', 'home_office', 'table', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to the living room and sit on the couch. Turn on television using remote control. Use remote control buttons to navigate to desired channel. Watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'hanger', 'floor', 'ceiling', 'cleaning_solution', 'mail', 'computer', 'toaster', 'mat', 'plate', 'faucet', 'brush', 'oil', 'candle', 'broom', 'wall', 'tape', 'bed', 'coffe_maker', 'ice', 'rag', 'console', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'colander']\nTask:Clean sink\nresponse:\n", "output": {"thoughts": "Squirt soap into the sink. Get a wet rag. Wipe out the sink. Rinse out the sink and the rag. Hang the rag out to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <cleaning_solution>", "[WALK] <sink>", "[POUR] <cleaning_solution> <sink>", "[PUTOBJBACK] <cleaning_solution>", "[WIPE] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <sink>", "[RINSE] <rag>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'dry_pasta', 'food_salt', 'chair', 'drawing', 'bills', 'hanger', 'floor', 'phone', 'ceiling', 'toaster', 'plate', 'wall', 'bench', 'powersocket', 'tape', 'cup', 'bed', 'check', 'food_food', 'tray', 'doorjamb', 'pencil', 'dishrack', 'dresser', 'cupboard', 'door', 'table', 'light', 'laser_pointer']\nTask:Eat\nresponse:\n{\n    \"thoughts\": Serve food on dinning table in appropriate utensils, sit on the chair, eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <plate>', '[PUTBACK] <food_food> <plate>', '[GRAB] <food_food>', '[PUTOBJBACK] <food_food>', '[FIND] <fork>', '[GRAB] <fork>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'detergent', 'drawing', 'floor', 'phone', 'bedroom', 'keyboard', 'mat', 'tvstand', 'stereo', 'clothes_socks', 'candle', 'towel', 'dining_room', 'wall', 'nightstand', 'powersocket', 'bed', 'check', 'food_food', 'razor', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'door', 'sink', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Turn on radio, put on favorite channel, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'bathroom', 'ceilingfan', 'milk', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'shoe-shine_kit', 'couch', 'food_jam', 'microphone', 'wall', 'nightstand', 'oven', 'cutting_board', 'television', 'curtain', 'pencil', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": We sit all around the television and watch a program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'walllamp', 'milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'candle', 'pillow', 'wall', 'cup', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'bathroom_cabinet', 'teeth']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTBACK] <remote_control> <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'scissors', 'microwave', 'jelly', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'shampoo', 'ceiling', 'keyboard', 'mail', 'brush', 'toaster', 'stovefan', 'plate', 'picture', 'oil', 'form', 'bowl', 'photoframe', 'newspaper', 'bed', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'home_office']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to living room. Find book. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'dog', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'laptop', 'mat', 'mousepad', 'wall', 'oven', 'love_seat', 'bed', 'coffe_maker', 'food_food', 'doorjamb', 'centerpiece', 'coffee_pot', 'mouse', 'pencil', 'hairbrush', 'controller', 'door', 'dishwasher', 'electrical_outlet']\nTask:Write book\nresponse:\n", "output": {"thoughts": "Writing a book", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <electrical_outlet>", "[PLUGIN] <laptop>", "[WALK] <love_seat>", "[SIT] <love_seat>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'computer', 'mail', 'table_cloth', 'mat', 'tvstand', 'plate', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'curtain', 'after_shave', 'doorjamb', 'cards', 'mouse', 'shoes', 'sink']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I sit in the sofa and I relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'spectacles', 'food_ice_cream', 'crayon', 'microwave', 'toilet', 'band-aids', 'bathroom', 'needle', 'bathroom_counter', 'man', 'alcohol', 'mat', 'plate', 'stereo', 'faucet', 'mousepad', 'couch', 'pillow', 'dining_room', 'light_bulb', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'home_office', 'table', 'board_game']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "When someone knocks I n the door walk through the hallway and greet them.", "actions": ["[RUN] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'food_salt', 'microwave', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'keyboard', 'food_cake', 'purse', 'shoe-shine_kit', 'faucet', 'wall', 'nightstand', 'powersocket', 'photoframe', 'check', 'coffe_maker', 'doorjamb', 'hair', 'sink', 'bathroom_cabinet', 'table']\nTask:Clean sink\nresponse:\n{\n    \"thoughts\": Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[WIPE] <faucet>', '[WIPE] <sink>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'toaster', 'closetdrawer', 'couch', 'food_jam', 'pillow', 'wall', 'cup', 'photoframe', 'coffee_table', 'novel', 'television', 'doorjamb', 'shoes', 'comb', 'dresser', 'instrument_violin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <coffee_table>", "[PUTBACK] <remote_control> <coffee_table>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'ground_coffee', 'bathroom', 'detergent', 'bathroom_counter', 'bedroom', 'bookshelf', 'button', 'toaster', 'mat', 'ceilinglamp', 'window', 'mousepad', 'couch', 'dining_room', 'photoframe', 'bed', 'novel', 'coffe_maker', 'food_food', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'sink', 'dishwasher', 'clothes_skirt', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to living room , find phone ,pick up phone 3,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>']\n}\n", "input": "\nScene_Objects:['sponge', 'chair', 'detergent', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'toaster', 'mat', 'plate', 'faucet', 'ceilinglamp', 'dish_soap', 'wall', 'food_cheese', 'light_bulb', 'novel', 'coffe_maker', 'food_food', 'television', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'shoes', 'dishrack', 'dresser', 'cupboard', 'sink', 'home_office']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Put all dirty dishes on one side. Take sponge and cover it in soap. Soap each dish. Rinse each dish with water. Put on drying rack.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[RINSE] <sponge>", "[SQUEEZE] <sponge>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <sponge>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'crayon', 'chair', 'cat', 'dough', 'milk', 'hanger', 'floor', 'keyboard', 'mail', 'table_cloth', 'stovefan', 'plate', 'faucet', 'tvstand', 'window', 'couch', 'form', 'dining_room', 'cutting_board', 'photoframe', 'check', 'wine', 'dresser', 'hairbrush', 'pajamas', 'conditioner', 'home_office', 'bathroom_cabinet', 'table']\nTask:Give milk to cat\nresponse:\n{\n    \"thoughts\": Grab milk and pour it into the plate so that the cat can drink,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[POINTAT] <cat>', '[PULL] <cat>', '[FIND] <milk>', '[GRAB] <milk>', '[FIND] <plate>', '[POUR] <milk> <plate>', '[TOUCH] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'drawing', 'floor', 'ceiling', 'bedroom', 'alcohol', 'toaster', 'stovefan', 'mat', 'shower', 'faucet', 'deck_of_cards', 'shoe-shine_kit', 'pillow', 'wall', 'oven', 'cup', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'table', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Turn on the faucet.  Place glass under tap and fill with water.  Turn off faucet.  Drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'plate', 'mousepad', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'wine', 'curtain', 'doorjamb', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'chair', 'freezer', 'drawing', 'orchid', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'ceilinglamp', 'plate', 'mat', 'stereo', 'shower', 'couch', 'pillow', 'food_kiwi', 'wall', 'pasta', 'bed', 'novel', 'food_food', 'check', 'television', 'console', 'pencil', 'kitchen_counter', 'cupboard', 'table']\nTask:Cut bread\nresponse:\n", "output": {"thoughts": "I get a slice of bread, grab a knife, and then cut the piece of bread in half.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <kitchen_cabinet>", "[FIND] <table>", "[PUTBACK] <plate> <table>", "[PUTBACK] <knife> <table>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[PUTBACK] <food_bread> <plate>", "[GRAB] <knife>", "[CUT] <food_bread>", "[PUTOBJBACK] <knife>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'nail_polish', 'trashcan', 'drawing', 'orchid', 'floor', 'desk', 'computer', 'food_bacon', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'bench', 'coffe_maker', 'television', 'curtain', 'mouse', 'bathtub', 'home_office', 'table', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'juice', 'microwave', 'chair', 'freezer', 'detergent', 'drawing', 'orchid', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'shower', 'mat', 'ceilinglamp', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'newspaper', 'coffe_maker', 'television', 'ice', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'cupboard', 'table', 'dishwasher', 'creditcard']\nTask:Grab some juice\nresponse:\n", "output": {"thoughts": "I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTOBJBACK] <juice>", "[CLOSE] <freezer>", "[FIND] <table>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'freezer', 'detergent', 'tea_bag', 'food_bacon', 'mop_bucket', 'stereo', 'iron', 'couch', 'pasta', 'powersocket', 'oven', 'wall_clock', 'photoframe', 'coffe_maker', 'mouthwash', 'television', 'food_rice', 'tray', 'after_shave', 'towel_rack', 'kitchen_counter', 'cd', 'dishrack', 'comb', 'dresser', 'clothes_skirt', 'creditcard']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to the radio.  I switch on the radio and listen,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['mop', 'chair', 'jelly', 'walllamp', 'drawing', 'wallshelf', 'food_egg', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'food_rice', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'coffee']\nTask:Watch  horror  movie\nresponse:\n", "output": {"thoughts": "I would  find something  scary  to  watch on dvd", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_ice_cream', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'longboard', 'floor', 'bookshelf', 'computer', 'faucet', 'balanceball', 'window', 'iron', 'food_peanut_butter', 'food_jam', 'wall', 'nightstand', 'bench', 'powersocket', 'basket_for_clothes', 'novel', 'check', 'coffee_filter', 'shaving_cream', 'pencil', 'maindoor', 'sink', 'hairdryer']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter the Home office, then walk to the phone, pickup the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'bed', 'television', 'doorjamb', 'shaving_cream', 'mouse', 'coffee', 'dresser', 'hairbrush', 'clothes_scarf', 'soap', 'teeth']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sitting on the couch watching my favorite television show.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'tea', 'microwave', 'dog', 'freezer', 'walllamp', 'drawing', 'orchid', 'hanger', 'keyboard', 'button', 'ceilinglamp', 'stereo', 'homework', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'food_jam', 'wall', 'oven', 'food_cheese', 'keys', 'tray', 'lighter', 'mouse', 'kitchen_counter', 'cd', 'clothes_dress', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Petting the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'dough', 'milk', 'drawing', 'floor', 'ceiling', 'keyboard', 'food_bacon', 'toaster', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'bowl', 'wall', 'oven', 'cup', 'coffe_maker', 'check', 'television', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'light', 'dishwasher']\nTask:Get something to drink\nresponse:\n", "output": {"thoughts": "Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[CLOSE] <freezer>", "[POUR] <milk> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'beer', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'mat', 'stereo', 'couch', 'pillow', 'bowl', 'wall', 'nightstand', 'powersocket', 'cup', 'toilet_paper', 'television', 'after_shave', 'rag', 'doorjamb', 'coffee_pot', 'dishrack', 'instrument_guitar', 'table', 'soap', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'blow_dryer', 'walllamp', 'knife', 'orchid', 'cpuscreen', 'floor', 'computer', 'button', 'ceilinglamp', 'oil', 'mousepad', 'cup', 'bed', 'keys', 'novel', 'check', 'coffe_maker', 'toothbrush_holder', 'coffee_filter', 'food_food', 'curtain', 'pencil', 'shoes', 'controller', 'hairdryer', 'bathroom_cabinet', 'table', 'board_game']\nTask:Get ready to leave\nresponse:\n", "output": {"thoughts": "Get your jacket and shoes on, locate the keys off the side table, walk out the front door and secure building with keys when you leave", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>", "[FIND] <keys>", "[GRAB] <keys>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'trashcan', 'dough', 'orchid', 'knife', 'clothes_underwear', 'phone', 'tea_bag', 'envelope', 'mat', 'faucet', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bench', 'powersocket', 'basket_for_clothes', 'food_food', 'food_rice', 'tray', 'curtain', 'after_shave', 'doorjamb', 'pencil', 'laundry_detergent', 'bathroom_cabinet', 'table', 'light', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'character', 'tea_bag', 'computer', 'table_cloth', 'toaster', 'mat', 'tvstand', 'stovefan', 'electric_shaver', 'couch', 'pillow', 'wall', 'basket_for_clothes', 'newspaper', 'coffee_table', 'photoframe', 'video_game_controller', 'television', 'doorjamb', 'dresser', 'cupboard', 'dishwasher']\nTask:Read magazine\nresponse:\n", "output": {"thoughts": "Get magazine off coffee table. Sit in arm chair. Open magazine to first page. Read until finished.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <newspaper>", "[PUTBACK] <newspaper> <coffee_table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'band-aids', 'walllamp', 'detergent', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'video_game_console', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'plate', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'photoframe', 'video_game_controller', 'television', 'curtain', 'dresser', 'table', 'dishwasher']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on game.  Use controller,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[PLUGIN] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'laptop', 'mat', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'cup', 'photoframe', 'television', 'doorjamb', 'mouse', 'door', 'creditcard']\nTask:Play on laptop\nresponse:\n", "output": {"thoughts": "Sit down on the couch. Place laptop on lap. Open laptop. Turn on laptop. Choose selected program. Play selected program. Close game. Close laptop.", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <couch>", "[SIT] <couch>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <laptop>", "[PUTOBJBACK] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'floor', 'character', 'desk', 'bedroom', 'keyboard', 'thread', 'toaster', 'plate', 'couch', 'wall', 'light_bulb', 'novel', 'coffe_maker', 'food_rice', 'tray', 'food_onion', 'after_shave', 'curtain', 'kettle', 'kitchen_counter', 'shoes', 'cupboard', 'conditioner', 'door', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office']\nTask:Check email\nresponse:\n", "output": {"thoughts": "Turn on the computer.  Wait for it to load.  Get online.  Go to the email service.  Open the email.  Read the email.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mail>", "[GRAB] <mail>", "[READ] <mail>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'coffe_maker', 'television', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'home_office', 'table', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the living room and find the book you want to read. Take it and then go to the couch. Sit on it, open the book and start reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'juice', 'nail_polish', 'wallshelf', 'floor', 'ceiling', 'face', 'mat', 'iron', 'shower', 'stereo', 'clothes_socks', 'food_jam', 'wall_clock', 'light_bulb', 'food_food', 'razor', 'mirror', 'food_rice', 'curtain', 'centerpiece', 'towel_rack', 'hair', 'clothes_dress', 'napkin', 'bathroom_cabinet', 'home_office', 'table', 'soap', 'shoe_rack']\nTask:Check appearance in mirror\nresponse:\n", "output": {"thoughts": "Walk to full-length mirror. Look in mirror. Make sure everything about your appearance is adequate. Adjust clothes as needed.", "actions": ["[WALK] <bedroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>", "[FIND] <face>", "[TURNTO] <face>", "[LOOKAT] <face>", "[FIND] <hair>", "[TURNTO] <hair>", "[LOOKAT] <hair>", "[FIND] <clothes_dress>", "[TURNTO] <clothes_dress>", "[LOOKAT] <clothes_dress>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'cpuscreen', 'hanger', 'floor', 'bookshelf', 'bookmark', 'stovefan', 'mat', 'faucet', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'cutting_board', 'cup', 'newspaper', 'novel', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'hair', 'clothes_dress', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room and relax on the sofa with a book,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[DROP] <bookmark>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['wallshelf', 'phone', 'dvd_player', 'character', 'computer', 'ceilinglamp', 'foundation', 'mousepad', 'bowl', 'wall', 'cutting_board', 'box', 'cup', 'food_cheese', 'bed', 'photoframe', 'coffe_maker', 'food_food', 'doorjamb', 'centerpiece', 'mouse', 'shoes', 'comb', 'dresser', 'conditioner', 'clothes_dress', 'hairdryer', 'spoon', 'bathroom_cabinet', 'light']\nTask:Leave home\nresponse:\n", "output": {"thoughts": "Put on shoes. Put on coat. Open door and exit through it. Lock door behind you.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'microwave', 'freezer', 'walllamp', 'detergent', 'dough', 'floor', 'phone', 'ceiling', 'tea_bag', 'bookshelf', 'keyboard', 'toaster', 'laptop', 'tvstand', 'faucet', 'mousepad', 'dining_room', 'wall', 'box', 'cup', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'dishwasher']\nTask:Make popcorn\nresponse:\n{\n    \"thoughts\": Lay popcorn bag flat inside microwave and press popcorn button to start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <microwave>', '[TURNTO] <microwave>', '[OPEN] <microwave>', '[PUTIN] <food_food> <microwave>', '[CLOSE] <microwave>', '[SWITCHON] <microwave>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_pizza', 'jelly', 'freezer', 'bathroom', 'detergent', 'food_butter', 'food_carrot', 'cpuscreen', 'bedroom', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'mousepad', 'food_noodles', 'food_orange', 'nightstand', 'cup', 'check', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'napkin', 'spoon', 'table', 'home_office', 'soap']\nTask:Open door\nresponse:\n", "output": {"thoughts": "Open the door when leaving or coming in.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'cat', 'floor', 'character', 'stamp', 'table_cloth', 'piano_bench', 'mat', 'window', 'couch', 'food_jam', 'pillow', 'wall', 'nightstand', 'bench', 'toy', 'cup', 'mouthwash', 'television', 'coffee_filter', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'clothes_dress', 'door', 'food_steak', 'bathtub', 'bag', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I see my cat on the couch so I walk over, sit down and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'food_food', 'television', 'doorjamb', 'mouse', 'pencil', 'dresser', 'door', 'shoe_rack']\nTask:Put in chair\nresponse:\n", "output": {"thoughts": "Take a comfortable chair. Place it in the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[GRAB] <chair>", "[WALK] <home_office>", "[WALK] <floor>", "[FIND] <floor>", "[PUTBACK] <chair> <floor>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'ground_coffee', 'walllamp', 'trashcan', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'toaster', 'shoe-shine_kit', 'faucet', 'clothes_pants', 'dining_room', 'wall', 'photoframe', 'clothes_gloves', 'coffe_maker', 'coffee_filter', 'centerpiece', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'clothes_scarf', 'home_office', 'table', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open the coffee maker and put a filer, grab ground coffee and put it in the coffee maker, then close the coffee maker and switch it on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[FIND] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'cup', 'newspaper', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Start computer\nresponse:\n", "output": {"thoughts": "Sit in desk chair, press power button, let computer power on", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'toilet', 'chair', 'walllamp', 'bathroom', 'orchid', 'floor', 'ceiling', 'bookshelf', 'table_cloth', 'bookmark', 'ceilinglamp', 'mat', 'stereo', 'shower', 'floor_lamp', 'food_noodles', 'clothes_hat', 'dining_room', 'wall', 'television', 'food_donut', 'doorjamb', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'bathroom_cabinet', 'light', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to wall, find light switch, turn on lights,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[WALK] <bedroom>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['food_bread', 'freezer', 'walllamp', 'drawing', 'needle', 'floor', 'ceiling', 'bookshelf', 'toaster', 'mat', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'food_kiwi', 'wall', 'cup', 'coffe_maker', 'television', 'food_donut', 'wine', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'food_steak', 'sink', 'dishwasher', 'electrical_outlet']\nTask:Toast bread\nresponse:\n", "output": {"thoughts": "Place two slices of bread in the toaster. Select medium toast on toaster.  Push down button to begin toasting.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[WALK] <electrical_outlet>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <electrical_outlet>", "[TURNTO] <toaster>", "[LOOKAT] <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'jelly', 'bathroom', 'knife', 'dvd_player', 'character', 'bedroom', 'thread', 'computer', 'toaster', 'shoe-shine_kit', 'deck_of_cards', 'shower', 'electric_shaver', 'homework', 'food_peanut_butter', 'form', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'novel', 'light_bulb', 'food_food', 'curtain', 'coffee_pot', 'food_steak', 'door', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the radio. Turn on the radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'food_butter', 'drawing', 'orchid', 'bills', 'wallshelf', 'cpuscreen', 'floor', 'shampoo', 'character', 'thread', 'alcohol', 'piano_bench', 'brush', 'ceilinglamp', 'mat', 'iron', 'dining_room', 'wall', 'wall_clock', 'pasta', 'newspaper', 'food_food', 'television', 'console', 'kitchen_counter', 'pajamas', 'table', 'shoe_rack']\nTask:Put down bags\nresponse:\n", "output": {"thoughts": "When walking into the house, putting down the things i am carrying", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <table>", "[PUTBACK] <food_food> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'trashcan', 'toothbrush', 'drawing', 'orchid', 'needle', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'bookshelf', 'remote_control', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'wall', 'powersocket', 'television', 'food_rice', 'doorjamb', 'pencil', 'chessboard', 'door', 'hairdryer', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'beer', 'detergent', 'food_butter', 'orchid', 'hanger', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'closetdrawer', 'broom', 'wall', 'nightstand', 'toy', 'newspaper', 'toilet_paper', 'television', 'wine', 'doorjamb', 'kitchen_counter', 'shoes', 'chessboard', 'dresser', 'door', 'dishwasher', 'instrument_violin', 'laser_pointer']\nTask:Put away shoes\nresponse:\n", "output": {"thoughts": "Pick up shoes from entry hall. Open entry way closet. Put shoes in closet. Close closet door.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[FIND] <dresser>", "[OPEN] <dresser>", "[PUTIN] <shoes> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blow_dryer', 'bathroom', 'food_carrot', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stamp', 'bedroom', 'food_cake', 'computer', 'shower', 'picture', 'oven', 'basket_for_clothes', 'photoframe', 'food_food', 'check', 'towel_rack', 'mouse', 'pencil', 'shoes', 'comb', 'hairbrush', 'cupboard', 'laundry_detergent', 'sink', 'napkin', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Pick up my dirty clothes, load the washer, add detergent, turn on washer,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[RUN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'food_food', 'television', 'doorjamb', 'mouse', 'cupboard', 'conditioner', 'door', 'bathroom_cabinet', 'light', 'table']\nTask:Look at painting\nresponse:\n", "output": {"thoughts": "Walk to the painting.  Look up at the painting.  Admire it on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'toilet', 'dough', 'toothbrush', 'drawing', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'stereo', 'shower', 'window', 'oil', 'couch', 'bowl', 'wall', 'oven', 'toy', 'cup', 'bed', 'food_food', 'television', 'wooden_spoon', 'towel_rack', 'comb', 'cupboard', 'table', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Find couch. Sit on couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'freezer', 'walllamp', 'bathroom', 'dough', 'knife', 'orchid', 'thread', 'computer', 'ceilinglamp', 'deck_of_cards', 'shower', 'clothes_socks', 'food_noodles', 'dustpan', 'nightstand', 'photoframe', 'light_bulb', 'television', 'coffee_filter', 'mirror', 'curtain', 'doorjamb', 'laundry_detergent', 'sink', 'bag', 'table', 'soap', 'clothes_shirt', 'creditcard']\nTask:Look at mirror\nresponse:\n", "output": {"thoughts": "I look at my reflection to check my cleanliness.", "actions": ["[WALK] <bathroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'bathroom', 'trashcan', 'wallshelf', 'clothes_underwear', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'faucet', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall_clock', 'newspaper', 'coffe_maker', 'television', 'curtain', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'bathtub', 'sink', 'clothes_scarf', 'light']\nTask:Watch fly\nresponse:\n{\n    \"thoughts\": Look around for a fly, often on a wall or ceiling.  Follow it's movements.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <fly>', '[FIND] <fly>', '[TURNTO] <fly>', '[LOOKAT] <fly>']\n}\n", "input": "\nScene_Objects:['glue', 'chair', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'mat', 'mousepad', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'food_food', 'coffe_maker', 'doorjamb', 'pencil', 'kitchen_counter', 'cd', 'shoes', 'hair', 'clothes_dress', 'table']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'keyboard', 'mat', 'ceilinglamp', 'shower', 'oil', 'food_peanut_butter', 'food_orange', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'pencil', 'dresser', 'door', 'napkin', 'light', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": put away groceries in fridge. open the door, put on shelf, close the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'drawing', 'wallshelf', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'mat', 'closetdrawer', 'couch', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'powersocket', 'oven', 'basket_for_clothes', 'photoframe', 'mouthwash', 'doorjamb', 'kitchen_counter', 'shoes', 'dresser', 'door', 'sink', 'light', 'clothes_shirt', 'teeth', 'creditcard']\nTask:Organize closet\nresponse:\n", "output": {"thoughts": "Organize the hanging clothes in my closet by color and type and get rid of clothes I don't wear.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PULL] <clothes_pants>", "[PUTIN] <clothes_pants> <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PULL] <clothes_pants>", "[PUTIN] <clothes_pants> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUSH] <clothes_shirt>", "[PUTIN] <clothes_shirt> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUSH] <clothes_shirt>", "[PUTIN] <clothes_shirt> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[CLOSE] <dresser>", "[WALK] <basket_for_clothes>", "[PUTBACK] <clothes_shirt> <basket_for_clothes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'microwave', 'toilet', 'jelly', 'freezer', 'band-aids', 'drawing', 'knifeblock', 'clothes_underwear', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'powersocket', 'cup', 'basket_for_clothes', 'photoframe', 'food_food', 'water', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Get drink\nresponse:\n{\n    \"thoughts\": I go to the fridge and open the door. I look for the water pitcher, and take it out, making sure to close the fridge door behind me. Then I take a pint glass out of the upper cabinet over the sink and use the water pitcher to pour a full glass of water. Then I return the pitcher to the fridge and shut the door again, and take my glass of water out of the kitche.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[OPEN] <freezer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'freezer', 'walllamp', 'detergent', 'bills', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'shower', 'faucet', 'closetdrawer', 'microphone', 'dining_room', 'pasta', 'mouthwash', 'mouse', 'pencil', 'dresser', 'pajamas', 'controller', 'sink', 'napkin', 'home_office', 'dishwasher', 'soap']\nTask:Switch on lamp\nresponse:\n", "output": {"thoughts": "Press the button so the light comes on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'drawing', 'cpuscreen', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'alcohol', 'laptop', 'plate', 'faucet', 'homework', 'mousepad', 'dining_room', 'check', 'food_food', 'television', 'tray', 'wooden_spoon', 'mouse', 'shoes', 'conditioner', 'maindoor', 'napkin', 'home_office', 'soap', 'instrument_violin']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I turn on computer. I use keyboard to type. I open browser. I search.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'cleaning_solution', 'bookshelf', 'table_cloth', 'mat', 'ceilinglamp', 'tvstand', 'foundation', 'closetdrawer', 'couch', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'television', 'rag', 'doorjamb', 'dresser', 'door', 'home_office', 'light']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Gather wood cleaner and cleaning rag. Wipe baseboards and window sills with dry cloth, then spray with cleaner and buff", "actions": ["[WALK] <bedroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <wall>", "[WALK] <window>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'homework', 'mousepad', 'clothes_pants', 'wall', 'nightstand', 'photoframe', 'check', 'curtain', 'doorjamb', 'mouse', 'light', 'dirt']\nTask:Work\nresponse:\n{\n    \"thoughts\": Find a PC and do homework,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[FIND] <mouse>', '[FIND] <check>', '[FIND] <homework>', '[GRAB] <homework>', '[FIND] <homework>', '[READ] <homework>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['glue', 'sponge', 'measuring_cup', 'toilet', 'food_butter', 'cpuscreen', 'bathroom_counter', 'ceiling', 'phone', 'character', 'face', 'food_bacon', 'shoe-shine_kit', 'closetdrawer', 'candle', 'food_jam', 'towel', 'pillow', 'nightstand', 'oven', 'cup', 'mouthwash', 'television', 'food_rice', 'after_shave', 'diary', 'headset', 'hair', 'door', 'light']\nTask:Rain welcome\nresponse:\n", "output": {"thoughts": "The robot would greet me and then hand me a towel.  It would make an appropriate quip about the weather while I dried off.  It would receive the wet towel when I was done.", "actions": ["[WALK] <bedroom>", "[WALK] <towel>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <face>", "[WIPE] <face>", "[FIND] <hair>", "[WIPE] <hair>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_bread', 'chair', 'freezer', 'walllamp', 'detergent', 'trashcan', 'drawing', 'desk', 'keyboard', 'bookshelf', 'bookmark', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'pillow', 'bowl', 'wall', 'box', 'coffe_maker', 'razor', 'television', 'curtain', 'doorjamb', 'towel_rack', 'door', 'sink', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Pour water into a glass and drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['sheets', 'crayon', 'chair', 'walllamp', 'bathroom', 'ceiling', 'bedroom', 'keyboard', 'bookshelf', 'clothes_jacket', 'toaster', 'deck_of_cards', 'faucet', 'mousepad', 'microphone', 'wall', 'tape', 'novel', 'food_food', 'check', 'food_donut', 'tray', 'towel_rack', 'dresser', 'pajamas', 'clothes_dress', 'hairdryer', 'bathroom_cabinet', 'home_office', 'table']\nTask:Grab things\nresponse:\n", "output": {"thoughts": "The robot will need some kind of hands or grabbing device to transport objects.", "actions": ["[WALK] <novel>", "[TURNTO] <novel>", "[GRAB] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'toilet', 'band-aids', 'food_carrot', 'dough', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'folder', 'bedroom', 'shoe-shine_kit', 'closetdrawer', 'food_orange', 'nightstand', 'wall', 'powersocket', 'coin', 'shaving_cream', 'pencil', 'cupboard', 'sink', 'napkin', 'instrument_guitar', 'bathroom_cabinet', 'soap', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'food_salt', 'measuring_cup', 'microwave', 'chair', 'orchid', 'knife', 'food_egg', 'wallshelf', 'desk', 'alcohol', 'computer', 'bookmark', 'shower', 'stereo', 'mousepad', 'dining_room', 'oven', 'powersocket', 'photoframe', 'mouthwash', 'tray', 'curtain', 'mouse', 'vacuum_cleaner', 'pajamas', 'sink', 'dishwasher', 'dirt']\nTask:Close door\nresponse:\n", "output": {"thoughts": "Push the door shut until it latches.", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'iron', 'mousepad', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'video_game_controller', 'television', 'doorjamb', 'kettle', 'mouse', 'door', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on Computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['juice', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'thread', 'computer', 'stovefan', 'mat', 'tvstand', 'toaster', 'couch', 'pillow', 'wall', 'photoframe', 'check', 'television', 'doorjamb', 'laundry_detergent', 'conditioner', 'sink', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Do homework\nresponse:\n", "output": {"thoughts": "Sit on the couch in the Living room and pick u some papers", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <check>", "[GRAB] <check>", "[DROP] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'toilet', 'dog', 'detergent', 'orchid', 'wallshelf', 'food_egg', 'cpuscreen', 'phone', 'stamp', 'bedroom', 'bookshelf', 'bookmark', 'shower', 'mousepad', 'food_jam', 'towel', 'pasta', 'bed', 'novel', 'food_food', 'curtain', 'centerpiece', 'shaving_cream', 'dresser', 'cupboard', 'table', 'creditcard']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the entrance hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'walllamp', 'food_butter', 'cpuscreen', 'phone', 'folder', 'bedroom', 'computer', 'stovefan', 'tvstand', 'plate', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'check', 'razor', 'television', 'tray', 'curtain', 'diary', 'shaving_cream', 'door', 'soap']\nTask:Open curtains\nresponse:\n", "output": {"thoughts": "Gently pull the curtains open, from the middle to the sides.", "actions": ["[WALK] <home_office>", "[WALK] <curtain>", "[FIND] <curtain>", "[PULL] <curtain>", "[FIND] <curtain>", "[PULL] <curtain>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'freezer', 'cat', 'food_carrot', 'dough', 'orchid', 'needle', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'ceilinglamp', 'mat', 'stereo', 'couch', 'dustpan', 'pillow', 'wall', 'powersocket', 'light_bulb', 'food_food', 'coffe_maker', 'video_game_controller', 'oven_mitts', 'food_rice', 'vacuum_cleaner', 'dresser', 'door', 'hairdryer']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I leave the bags with groceries near the fridge, open the door and put the groceries in the shelves.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'laptop', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'tape', 'photoframe', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'light']\nTask:Use laptop\nresponse:\n", "output": {"thoughts": "Sit down on a couch and perform tasks such as navigating the internet on a laptop", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <keyboard>", "[LOOKAT] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'faucet', 'shower', 'tooth_paste', 'towel', 'wall', 'wall_clock', 'oven', 'tape', 'photoframe', 'clothes_gloves', 'light_bulb', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'light', 'vase', 'soap', 'home_office', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, open the faucet rinse the hands, wash the hands with soap and rinse again, then wipe with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'tea', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'dresser', 'door', 'home_office']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Keep my keyboard vacuumed and free of debris.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'bathroom', 'cat', 'trashcan', 'drawing', 'wallshelf', 'floor', 'desk', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'bench', 'cup', 'food_food', 'television', 'curtain', 'towel_rack', 'mouse', 'sink', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'drawing', 'clothes_underwear', 'hanger', 'bathroom_counter', 'bedroom', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mat', 'mousepad', 'couch', 'pillow', 'dining_room', 'food_food', 'check', 'coffe_maker', 'mirror', 'ice', 'wooden_spoon', 'rag', 'cupboard', 'door', 'sink', 'table']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray glass cleaner on mirror. Wipe with a clean cloth.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[WALK] <mirror>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'orchid', 'needle', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to my desk. I turn on the computer. I open a web page and look at it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'freezer', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'button', 'mat', 'shoe-shine_kit', 'pillow', 'wall', 'oven', 'tape', 'bed', 'photoframe', 'novel', 'food_food', 'curtain', 'coffee_pot', 'mouse', 'pencil', 'conditioner', 'sink', 'coffee']\nTask:Read to child\nresponse:\n", "output": {"thoughts": "Read a book out loud to a child", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'phone', 'remote_control', 'food_bacon', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'powersocket', 'oven', 'cup', 'bed', 'coffe_maker', 'television', 'curtain', 'lighter', 'doorjamb', 'shaving_cream', 'headset', 'cupboard', 'table', 'clothes_shirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": We recently got a 43 inch TV and have enjoyed watching some of our favorite movies and TV shows. We use a remote to do so.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['crayon', 'chair', 'detergent', 'cat', 'toothbrush', 'phone', 'character', 'purse', 'thread', 'computer', 'ceilinglamp', 'stereo', 'couch', 'food_peanut_butter', 'towel', 'wall', 'cup', 'food_cheese', 'photoframe', 'light_bulb', 'coffe_maker', 'television', 'towel_rack', 'kitchen_counter', 'dishrack', 'comb', 'dresser', 'door', 'sink', 'instrument_violin']\nTask:Keep cats out of room\nresponse:\n", "output": {"thoughts": "Stand in the doorway and create a loud noise so the cats will stay out.", "actions": ["[WALK] <bedroom>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'towel', 'wall', 'wall_clock', 'oven', 'television', 'tray', 'rag', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <bathroom_cabinet>', '[FIND] <shower>', '[RINSE] <shower>']\n}\n", "input": "\nScene_Objects:['sheets', 'cpuscreen', 'phone', 'desk', 'cleaning_solution', 'keyboard', 'purse', 'table_cloth', 'stovefan', 'tvstand', 'shower', 'stereo', 'food_noodles', 'towel', 'pillow', 'wall', 'wall_clock', 'oven', 'basket_for_clothes', 'photoframe', 'novel', 'food_rice', 'rag', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'napkin', 'table']\nTask:Clean  mirror\nresponse:\n", "output": {"thoughts": "GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'bathroom', 'trashcan', 'wallshelf', 'clothes_underwear', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'faucet', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall_clock', 'newspaper', 'coffe_maker', 'television', 'curtain', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'bathtub', 'sink', 'clothes_scarf', 'light']\nTask:Watch fly\nresponse:\n{\n    \"thoughts\": Look around for a fly, often on a wall or ceiling.  Follow it's movements.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <fly>', '[FIND] <fly>', '[TURNTO] <fly>', '[LOOKAT] <fly>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'light', 'table']\nTask:Straighten paintings on wall\nresponse:\n", "output": {"thoughts": "Go up to the painting and tilt it which ever way it is uneven until it looks straight.", "actions": ["[WALK] <home_office>", "[WALK] <wall>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'pot', 'microwave', 'toilet', 'band-aids', 'walllamp', 'trashcan', 'knifeblock', 'drawing', 'ceiling', 'desk', 'stovefan', 'faucet', 'mousepad', 'wall', 'nightstand', 'oven', 'check', 'food_food', 'water', 'tray', 'doorjamb', 'kitchen_counter', 'cd', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'soap', 'laser_pointer']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['blender', 'tea', 'microwave', 'jelly', 'band-aids', 'freezer', 'bathroom', 'needle', 'ceiling', 'character', 'keyboard', 'mop_bucket', 'ceilinglamp', 'mat', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'oven', 'photoframe', 'coffe_maker', 'check', 'television', 'cards', 'laundry_detergent', 'clothes_dress', 'home_office', 'table']\nTask:Put umbrella away\nresponse:\n", "output": {"thoughts": "Place the umbrella in the umbrella bucket point first.", "actions": ["[WALK] <home_office>", "[WALK] <mop_bucket>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": you go into your office, switch on the light, find chair at desk, sit in the chair, switch on the computer and work.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'tea_bag', 'mat', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'toy', 'photoframe', 'check', 'mouthwash', 'food_food', 'food_donut', 'wine', 'curtain', 'doorjamb', 'console', 'dresser', 'door', 'light']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Bend down and pick up the toys off of the floor.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[WALK] <dresser>", "[GRAB] <toy>", "[PUTIN] <toy> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'toothbrush', 'cpuscreen', 'clothes_underwear', 'wallshelf', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'computer', 'button', 'pillow', 'dining_room', 'bench', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'dresser', 'cupboard', 'pajamas', 'clothes_dress', 'sink', 'home_office', 'laser_pointer']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on Computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'walllamp', 'drawing', 'floor', 'desk', 'tea_bag', 'computer', 'clothes_jacket', 'toaster', 'mat', 'mousepad', 'closetdrawer', 'towel', 'tape', 'basket_for_clothes', 'bed', 'curtain', 'shaving_cream', 'towel_rack', 'mouse', 'kitchen_counter', 'chessboard', 'dresser', 'hair', 'laundry_detergent', 'bathroom_cabinet', 'light', 'board_game', 'dirt']\nTask:Open front door\nresponse:\n", "output": {"thoughts": "Walk over to the front door, turn the knob, and open the door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'broom', 'wall', 'pasta', 'powersocket', 'shelf', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the computer and turn it on. I look at the computer screen. I move and click the mouse to select buttons on the screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[PUSH] <mouse>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['spectacles', 'walllamp', 'nail_polish', 'milk', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'food_bacon', 'piano_bench', 'mat', 'tvstand', 'stereo', 'couch', 'clothes_hat', 'form', 'pillow', 'dining_room', 'wall', 'wall_clock', 'cup', 'photoframe', 'novel', 'coffe_maker', 'television', 'ice', 'doorjamb', 'mouse', 'scrabble']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on a couch, grab a book, open it, and scan the pages", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'food_butter', 'toothbrush', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'ceilinglamp', 'mousepad', 'closetdrawer', 'oil', 'microphone', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'powersocket', 'bed', 'photoframe', 'razor', 'television', 'doorjamb', 'kettle', 'mouse', 'dresser', 'door', 'light']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Hit the switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['beer', 'measuring_cup', 'orchid', 'hanger', 'floor', 'ceiling', 'phone', 'tea_bag', 'keyboard', 'purse', 'computer', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'novel', 'food_food', 'television', 'doorjamb', 'coffee_pot', 'vacuum_cleaner', 'dresser', 'hairbrush', 'door', 'napkin', 'home_office', 'board_game']\nTask:Sweep hallway please\nresponse:\n", "output": {"thoughts": "Vacuum all the way up and down the hallway.  Get all the dirt and crumbs out of carpet please.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'dog', 'band-aids', 'walllamp', 'ground_coffee', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'phone', 'food_cake', 'computer', 'toaster', 'mat', 'faucet', 'window', 'couch', 'candle', 'pillow', 'wall', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'table', 'light', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sponge', 'filing_cabinet', 'measuring_cup', 'bathroom', 'trashcan', 'food_carrot', 'cpuscreen', 'character', 'computer', 'floor_lamp', 'closetdrawer', 'couch', 'food_peanut_butter', 'food_kiwi', 'oven', 'cup', 'bed', 'video_game_controller', 'doorjamb', 'coffee_pot', 'paper_towel', 'dishrack', 'headset', 'hair', 'cupboard', 'spoon', 'bathroom_cabinet', 'table', 'tooth_paste']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Clean lamp shades in the living room", "actions": ["[WALK] <home_office>", "[WALK] <paper_towel>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WALK] <floor_lamp>", "[WIPE] <floor_lamp>", "[FIND] <trashcan>", "[PUTBACK] <paper_towel> <trashcan>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WALK] <floor_lamp>", "[WIPE] <floor_lamp>", "[PUTBACK] <paper_towel> <trashcan>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'jelly', 'freezer', 'chair', 'blow_dryer', 'bathroom', 'walllamp', 'knife', 'phone', 'keyboard', 'stovefan', 'homework', 'couch', 'food_jam', 'bench', 'oven', 'toy', 'box', 'food_cheese', 'oven_mitts', 'food_rice', 'tray', 'curtain', 'after_shave', 'dresser', 'clothes_dress', 'conditioner', 'hairdryer', 'vase', 'board_game']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my living room, find my cat, scrub it using my hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['spectacles', 'walllamp', 'bathroom', 'blow_dryer', 'orchid', 'hanger', 'floor', 'ceiling', 'character', 'computer', 'mat', 'mousepad', 'closetdrawer', 'food_orange', 'dining_room', 'wall', 'nightstand', 'basket_for_clothes', 'coffe_maker', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'light', 'clothes_skirt', 'laser_pointer']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug in the vacuum. Turn it on. Push it back and forth across the floor until all the debris is gone.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'nail_polish', 'trashcan', 'milk', 'drawing', 'wallshelf', 'bathroom_counter', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'computer', 'faucet', 'pillow', 'wall', 'photoframe', 'toilet_paper', 'check', 'coffe_maker', 'wine', 'after_shave', 'towel_rack', 'comb', 'controller', 'sink', 'bathroom_cabinet', 'light', 'table', 'instrument_violin', 'teeth']\nTask:Get toilet paper\nresponse:\n{\n    \"thoughts\": There are extra toilet papers in the bottom cabinet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <toilet_paper>', '[GRAB] <toilet_paper>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'bathroom', 'drawing', 'orchid', 'floor', 'bookshelf', 'mat', 'tvstand', 'shower', 'oil', 'closetdrawer', 'candle', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'cup', 'bed', 'photoframe', 'tray', 'doorjamb', 'pencil', 'cupboard', 'door', 'home_office', 'light', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Climb into bed, then read the novel", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'stereo', 'balanceball', 'window', 'clothes_socks', 'food_noodles', 'couch', 'wall', 'oven', 'tape', 'check', 'television', 'doorjamb', 'mouse', 'pencil', 'cd', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'light', 'table', 'instrument_violin']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I want to relax on the sofa to some relaxing music in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <stereo>', '[PLUGIN] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light']\nTask:Research\nresponse:\n", "output": {"thoughts": "I read different articles online to gain a better understanding of a particular business, industry, method, terminology, etc.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'food_ice_cream', 'walllamp', 'bathroom', 'ceilingfan', 'trashcan', 'woman', 'drawing', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'bedroom', 'man', 'ceilinglamp', 'window', 'couch', 'wall', 'bench', 'cup', 'newspaper', 'novel', 'check', 'television', 'towel_rack', 'clothes_dress', 'table', 'clothes_skirt', 'creditcard']\nTask:Hang with friends\nresponse:\n{\n    \"thoughts\": Sit with friends. Talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[TURNTO] <man>', '[LOOKAT] <man>', '[GREET] <man>', '[FIND] <woman>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <man>', '[LOOKAT] <man>', '[GREET] <man>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'door', 'teeth', 'printing_paper']\nTask:Print out papers\nresponse:\n", "output": {"thoughts": "Turn on printer and fill with blank paper. Use computer to click on printing option. Take printed page out of print.", "actions": ["[WALK] <home_office>", "[WALK] <fax_machine>", "[FIND] <fax_machine>", "[SWITCHON] <fax_machine>", "[FIND] <printing_paper>", "[GRAB] <printing_paper>", "[PUTBACK] <printing_paper> <fax_machine>", "[FIND] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[GRAB] <printing_paper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'crayon', 'dog', 'freezer', 'knife', 'orchid', 'bathroom_counter', 'character', 'bedroom', 'faucet', 'couch', 'food_orange', 'towel', 'nightstand', 'oven', 'powersocket', 'photoframe', 'check', 'video_game_controller', 'food_donut', 'food_rice', 'food_onion', 'diary', 'dresser', 'hairbrush', 'clothes_scarf', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the cat. Pet the cat with your hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'love_seat', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'clothes_scarf']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit and watch the television", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'crayon', 'microwave', 'drawing', 'bills', 'cpuscreen', 'longboard', 'floor', 'bookshelf', 'computer', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'window', 'oil', 'couch', 'food_jam', 'food_orange', 'wall', 'bench', 'mouthwash', 'television', 'cupboard', 'table', 'clothes_skirt', 'light', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'food_ice_cream', 'toilet', 'walllamp', 'detergent', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'tvstand', 'laptop', 'ceilinglamp', 'brush', 'couch', 'pillow', 'dining_room', 'wall', 'child', 'oven', 'photoframe', 'bed', 'television', 'doorjamb', 'mouse', 'dishwasher']\nTask:Let baby learn how to walk\nresponse:\n", "output": {"thoughts": "The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[GREET] <child>", "[FIND] <couch>", "[TURNTO] <couch>", "[POINTAT] <couch>", "[TURNTO] <child>", "[LOOKAT] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_pizza', 'filing_cabinet', 'tablelamp', 'freezer', 'nail_polish', 'cpuscreen', 'desk', 'bookshelf', 'keyboard', 'alcohol', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'wall', 'wall_clock', 'bed', 'coffe_maker', 'mouthwash', 'video_game_controller', 'television', 'razor', 'check', 'tray', 'curtain', 'cd', 'spoon', 'vase']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find the cat. Walk over to the cat. Touch the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'detergent', 'food_butter', 'toothbrush', 'drawing', 'floor', 'ceiling', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'dish_soap', 'wall', 'nightstand', 'cup', 'coffe_maker', 'coffee_filter', 'drinking_glass', 'tray', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'sink', 'coffee_cup', 'dishwasher', 'shoe_rack']\nTask:Clean dishes\nresponse:\n", "output": {"thoughts": "Wash dishes in the sink then place them in the dishwasher. Put detergent in the dishwasher and turn it on when full.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <drinking_glass>", "[WASH] <drinking_glass>", "[FIND] <coffee_cup>", "[WASH] <coffee_cup>", "[FIND] <plate>", "[WASH] <plate>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <plate>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <detergent>", "[GRAB] <detergent>", "[PUTBACK] <detergent> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'food_bacon', 'toaster', 'tvstand', 'window', 'oil', 'couch', 'wall', 'nightstand', 'newspaper', 'coffe_maker', 'television', 'coffee_filter', 'water', 'food_donut', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'dirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'ceilinglamp', 'floor_lamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'oven', 'photoframe', 'novel', 'television', 'doorjamb', 'dresser', 'cupboard', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'tea', 'chair', 'freezer', 'ground_coffee', 'drawing', 'orchid', 'floor', 'purse', 'faucet', 'balanceball', 'food_noodles', 'food_peanut_butter', 'wall', 'powersocket', 'toy', 'bed', 'food_food', 'check', 'food_donut', 'doorjamb', 'pencil', 'laundry_detergent', 'door', 'bag', 'napkin', 'light', 'table']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Grab toys off floor.  Put it in a container.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'nightstand', 'feet_both', 'photoframe', 'bed', 'television', 'coin', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'shoes', 'hairbrush', 'door']\nTask:Put on your shoes\nresponse:\n", "output": {"thoughts": "Sit on a chair with your shoes in front of you.  Put them on one at a time.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <feet_both>", "[PUTBACK] <shoes> <feet_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'jelly', 'freezer', 'bathroom', 'drawing', 'orchid', 'character', 'folder', 'keyboard', 'mop_bucket', 'stovefan', 'faucet', 'food_noodles', 'couch', 'clothes_hat', 'towel', 'nightstand', 'powersocket', 'rag', 'towel_rack', 'pencil', 'kitchen_counter', 'cupboard', 'pajamas', 'bathroom_cabinet', 'table', 'dishwasher', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and get the soap and wash my both hand using the soap and rinse my hands and use towel.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[FIND] <towel>', '[RINSE] <towel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'coffee_cup']\nTask:Type up document\nresponse:\n", "output": {"thoughts": "A person at a computer desk with coffee on the side by the keyboard typing up a document. The person looks relaxed and is typing while looking at the screen which shows a text editor and the content the person was typing.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[DRINK] <coffee_cup>", "[PUTOBJBACK] <coffee_cup>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'chair', 'freezer', 'drawing', 'knifeblock', 'cpuscreen', 'needle', 'floor', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'brush', 'mousepad', 'wall', 'nightstand', 'bed', 'photoframe', 'check', 'food_food', 'doorjamb', 'mouse', 'kitchen_counter', 'headset', 'door', 'table', 'light', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I'll walk to my bedroom, where I have my computer. I'll walk to my computer and sit down in front of it. I'll put on my headphones and unlock my computer. I'll then navigate to music stored locally or streamed through the web.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <headset>', '[GRAB] <headset>', '[PUTON] <headset>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'tablelamp', 'microwave', 'chair', 'freezer', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'stovefan', 'shower', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'cup', 'coffe_maker', 'tray', 'doorjamb', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'table', 'dishwasher', 'light']\nTask:Wipe down counter\nresponse:\n", "output": {"thoughts": "Clean counter", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <rag>", "[SQUEEZE] <rag>", "[SWITCHOFF] <faucet>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'food_bread', 'filing_cabinet', 'tablelamp', 'freezer', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'bills', 'floor', 'ceiling', 'character', 'food_cake', 'piano_bench', 'brush', 'ceilinglamp', 'mousepad', 'towel', 'wall', 'bed', 'food_food', 'food_donut', 'ice', 'curtain', 'kettle', 'dresser', 'light', 'soap', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Open the fridge, put the groceries in the fridge and close it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_cake', 'computer', 'food_bacon', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'deck_of_cards', 'mousepad', 'couch', 'wall', 'nightstand', 'newspaper', 'television', 'mouse', 'door', 'sink', 'home_office']\nTask:Read newspaper\nresponse:\n", "output": {"thoughts": "Sit in my chair at my desk. Open up newspaper. Read about current events.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[OPEN] <newspaper>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'toothbrush', 'cpuscreen', 'clothes_underwear', 'wallshelf', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'computer', 'button', 'pillow', 'dining_room', 'bench', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'dresser', 'cupboard', 'pajamas', 'clothes_dress', 'sink', 'home_office', 'laser_pointer']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on Computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'bathroom', 'milk', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'oil', 'pillow', 'wall', 'nightstand', 'pasta', 'newspaper', 'television', 'towel_rack', 'mouse', 'headset', 'chessboard', 'door', 'sink']\nTask:Read news\nresponse:\n", "output": {"thoughts": "Read news articles in a newspaper", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'mop', 'microwave', 'toilet', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'computer', 'piano_bench', 'ceilinglamp', 'plate', 'deck_of_cards', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'bed', 'novel', 'television', 'doorjamb', 'door', 'coffee_cup', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['jelly', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'keyboard', 'bookshelf', 'envelope', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'newspaper', 'photoframe', 'check', 'television', 'doorjamb', 'hairbrush', 'pajamas', 'table', 'clothes_shirt', 'dirt']\nTask:Read newspaper\nresponse:\n", "output": {"thoughts": "Read the newspaper on the couch", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'toilet', 'freezer', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'character', 'desk', 'food_cake', 'toaster', 'ceilinglamp', 'couch', 'pillow', 'powersocket', 'basket_for_clothes', 'cup', 'newspaper', 'food_food', 'check', 'oven_mitts', 'wooden_spoon', 'centerpiece', 'kitchen_counter', 'door', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office, find where the phone is and lift the reciever.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tea', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'clothes_underwear', 'food_egg', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'bookmark', 'mat', 'tvstand', 'couch', 'pillow', 'wall', 'cup', 'photoframe', 'bed', 'toilet_paper', 'television', 'doorjamb', 'dresser', 'light', 'creditcard']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up a book and open to the last page you were on. Read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I want to surf the internet and look around at funny pictures.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'conditioner', 'door', 'table']\nTask:Check email\nresponse:\n", "output": {"thoughts": "I log onto my computer and then go to the e-mail app, I click on each new email and read them", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mail>", "[GRAB] <mail>", "[READ] <mail>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'detergent', 'food_butter', 'dough', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'purse', 'faucet', 'stereo', 'picture', 'dustpan', 'bowl', 'dining_room', 'wall', 'oven', 'powersocket', 'photoframe', 'wine', 'tray', 'console', 'pencil', 'sink', 'bathroom_cabinet', 'dishwasher', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'woman', 'toothbrush', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'stamp', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'child', 'box', 'photoframe', 'television', 'oven_mitts', 'doorjamb', 'cd', 'dresser', 'table']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sitting and watching  a movie with family", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TURNTO] <child>", "[LOOKAT] <child>", "[GREET] <child>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'dustpan', 'dining_room', 'wall', 'nightstand', 'pasta', 'tape', 'photoframe', 'tray', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'bathroom_cabinet', 'light']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'freezer', 'orchid', 'floor', 'bedroom', 'keyboard', 'stovefan', 'stereo', 'shower', 'closetdrawer', 'couch', 'dining_room', 'tape', 'cup', 'bed', 'novel', 'television', 'toothbrush_holder', 'doorjamb', 'mouse', 'kitchen_counter', 'chessboard', 'cupboard', 'door', 'spoon', 'instrument_guitar', 'table', 'clothes_shirt', 'shoe_rack']\nTask:Look out window\nresponse:\n", "output": {"thoughts": "I head into my child's room and look out the window to view the neighborhood from a different area in the house.", "actions": ["[WALK] <bedroom>", "[WALK] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_bread', 'crayon', 'measuring_cup', 'chair', 'walllamp', 'trashcan', 'bedroom', 'bookshelf', 'purse', 'table_cloth', 'faucet', 'clothes_socks', 'pillow', 'wall', 'oven', 'newspaper', 'photoframe', 'coffe_maker', 'food_food', 'check', 'coffee_filter', 'shaving_cream', 'pencil', 'headset', 'dishrack', 'comb', 'dresser', 'spoon', 'instrument_guitar']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to living room, find lamp, switch on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['scissors', 'tablelamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'alcohol', 'mat', 'tvstand', 'stereo', 'closetdrawer', 'couch', 'clothes_pants', 'food_jam', 'pillow', 'microphone', 'dining_room', 'wall', 'powersocket', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'hairdryer', 'light', 'instrument_violin', 'clothes_shirt']\nTask:Getting  dresses\nresponse:\n", "output": {"thoughts": "Putting on undergarments  and clothes", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'food_ice_cream', 'blow_dryer', 'detergent', 'ceilingfan', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'shoe-shine_kit', 'couch', 'food_orange', 'wall', 'tape', 'television', 'curtain', 'kitchen_counter', 'conditioner', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": For relaxing walk to living room. Find the sofa to relax. Arrange the sofa to relax. Sit on the sofa to relaxing.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <conditioner>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'dishwasher']\nTask:Work\nresponse:\n", "output": {"thoughts": "Sit at my computer desk, Turn on Computer. Get online. Work.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'ceilingfan', 'drawing', 'needle', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'clothes_socks', 'couch', 'pillow', 'wall', 'powersocket', 'food_cheese', 'bed', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit down on the sofa. Pick up the remote control. Turn on the television with the remote control. Put down the remote control. Watch the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I sit at my desk and work on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'band-aids', 'trashcan', 'dough', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'dvd_player', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'mousepad', 'food_orange', 'wall', 'oven', 'keys', 'mouse', 'shoes', 'hair', 'dishwasher', 'vase', 'coffee']\nTask:Read book\nresponse:\n{\n    \"thoughts\": find a book to read. sit down. read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Pay bills\nresponse:\n", "output": {"thoughts": "Check my bills, on the computer, open the appropriate sites, pay my due bills", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'ground_coffee', 'nail_polish', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'toaster', 'ceilinglamp', 'mat', 'faucet', 'stovefan', 'mousepad', 'oil', 'wall', 'bed', 'mouse', 'cupboard', 'door', 'hairdryer', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'child', 'photoframe', 'food_food', 'television', 'after_shave', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn on tv.  Sit with kids.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[FIND] <child>", "[GREET] <child>", "[FIND] <child>", "[GREET] <child>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>", "[TURNTO] <television>", "[WATCH] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'coffee_table', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'centerpiece', 'door', 'bathroom_cabinet', 'table', 'light', 'board_game']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'computer', 'bookmark', 'stovefan', 'mat', 'tvstand', 'faucet', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'television', 'doorjamb', 'coffee_pot', 'dresser', 'bathroom_cabinet', 'dishwasher', 'laser_pointer', 'shoe_rack']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Bring book, lie on the sofa, read book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[LIE] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'detergent', 'drawing', 'floor', 'ceiling', 'stamp', 'bedroom', 'thread', 'computer', 'toaster', 'faucet', 'mousepad', 'food_peanut_butter', 'wall', 'powersocket', 'toy', 'cup', 'newspaper', 'doorjamb', 'chessboard', 'comb', 'door', 'sink', 'bag', 'light', 'board_game']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Enter inside kid's room, begin picking up toys from the floor and putting them in a toy box, leave once once.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'food_cake', 'thread', 'computer', 'laptop', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'pasta', 'television', 'tray', 'doorjamb', 'mouse', 'headset', 'chessboard', 'cupboard', 'door']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sit in my deck chair with my laptop, find a good movie on You Tube to watch, plug in earphones and listen to movie quietly so I don't disturb my upstairs neighbor.", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <headset>", "[GRAB] <headset>", "[PUTON] <headset>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_bread', 'filing_cabinet', 'tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'mousepad', 'food_noodles', 'pillow', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'doorjamb', 'towel_rack', 'mouse', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to home office ,find computer ,switch on computer and work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'crayon', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'door', 'clothes_scarf', 'light']\nTask:Re arrange office\nresponse:\n", "output": {"thoughts": "Walk in.Turn on light.Move furniture. Change the room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[PUSH] <desk>", "[WALK] <chair>", "[PUSH] <chair>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'tea', 'toilet', 'walllamp', 'drawing', 'orchid', 'floor', 'desk', 'bedroom', 'computer', 'ceilinglamp', 'tvstand', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'food_cheese', 'bed', 'food_donut', 'curtain', 'food_onion', 'mouse', 'pajamas', 'controller', 'table', 'dishwasher', 'light', 'board_game']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": lie in the sofa, relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'microwave', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'toaster', 'stovefan', 'deck_of_cards', 'shower', 'pillow', 'wall', 'powersocket', 'cup', 'coffe_maker', 'check', 'wine', 'food_onion', 'curtain', 'lighter', 'kitchen_counter', 'cupboard', 'hair', 'controller', 'table', 'dishwasher', 'colander']\nTask:Take dishes out of dishwasher\nresponse:\n", "output": {"thoughts": "Take clean dishes out of the dishwasher and put away", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Find the computer.  Turn on the computer.  Open the e-mail program.  Create a new message.  Write the message.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'measuring_cup', 'toilet', 'chair', 'cat', 'milk', 'orchid', 'phone', 'video_game_console', 'bookshelf', 'keyboard', 'purse', 'table_cloth', 'brush', 'mat', 'iron', 'couch', 'cup', 'basket_for_clothes', 'video_game_controller', 'oven_mitts', 'food_donut', 'doorjamb', 'console', 'dresser', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Playing video game\nresponse:\n", "output": {"thoughts": "Playing  video games  on various video game  systems such as wii u  and  xbox", "actions": ["[WALK] <home_office>", "[WALK] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[PUTOBJBACK] <video_game_controller>", "[SWITCHOFF] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_oatmeal', 'microwave', 'freezer', 'bathroom', 'nail_polish', 'dough', 'food_carrot', 'orchid', 'drawing', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'brush', 'mat', 'couch', 'pillow', 'food_cereal', 'wall', 'food_vegetable', 'food_food', 'television', 'console', 'cards', 'kitchen_counter', 'cupboard', 'hair', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_vegetable> <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTIN] <food_carrot> <freezer>', '[FIND] <food_cereal>', '[GRAB] <food_cereal>', '[PUTIN] <food_cereal> <freezer>', '[FIND] <food_oatmeal>', '[GRAB] <food_oatmeal>', '[PUTIN] <food_oatmeal> <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[FIND] <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'mop', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office']\nTask:Social media  checks\nresponse:\n", "output": {"thoughts": "Logging on to  things like  facebook  and  twitter", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'juice', 'microwave', 'toilet', 'chair', 'walllamp', 'dough', 'drawing', 'cpuscreen', 'floor', 'folder', 'desk', 'bookshelf', 'keyboard', 'computer', 'mail', 'shower', 'picture', 'mousepad', 'wall', 'nightstand', 'bench', 'oven', 'photoframe', 'novel', 'check', 'food_food', 'centerpiece', 'mouse']\nTask:Arrange folders\nresponse:\n{\n    \"thoughts\": Go through all my folders and re-arrange in alfabetaical order and put on computer desk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <folder>', '[GRAB] <folder>', '[OPEN] <desk>', '[PUTBACK] <folder> <desk>', '[FIND] <folder>', '[GRAB] <folder>', '[PUTBACK] <folder> <desk>', '[FIND] <folder>', '[GRAB] <folder>', '[PUTBACK] <folder> <desk>', '[FIND] <folder>', '[GRAB] <folder>', '[PUTBACK] <folder> <desk>', '[FIND] <folder>', '[GRAB] <folder>', '[PUTBACK] <folder> <desk>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'cloth_napkin', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'pencil', 'headset', 'dresser', 'door', 'sink']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe away any marks on the computer screen.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <cloth_napkin>", "[GRAB] <cloth_napkin>", "[WIPE] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'mop', 'chair', 'freezer', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'toaster', 'shower', 'window', 'food_orange', 'wall', 'oven', 'cup', 'photoframe', 'food_food', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'spoon', 'table', 'light', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'food_butter', 'toothbrush', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'ceilinglamp', 'mousepad', 'closetdrawer', 'oil', 'microphone', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'powersocket', 'bed', 'photoframe', 'razor', 'television', 'doorjamb', 'kettle', 'mouse', 'dresser', 'door', 'light']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Hit the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'sink', 'light', 'coffee']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room ,find television , switch on remote control,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[SWITCHON] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'microwave', 'chair', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'brush', 'mat', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'wall', 'nightstand', 'powersocket', 'food_food', 'television', 'curtain', 'lighter', 'mouse', 'door', 'clothes_scarf']\nTask:Enjoy view out window\nresponse:\n", "output": {"thoughts": "Sitting in teak deck chair, gaze out window at the view.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'ground_coffee', 'trashcan', 'desk', 'bedroom', 'purse', 'mop_bucket', 'shower', 'picture', 'homework', 'mousepad', 'form', 'wall', 'oven', 'bed', 'keys', 'video_game_controller', 'wine', 'food_onion', 'after_shave', 'console', 'mouse', 'cd', 'comb', 'dresser', 'door', 'clothes_skirt', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find the cat; pet the cat by touching it,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'food_pizza', 'juice', 'microwave', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'toaster', 'mat', 'stovefan', 'shower', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'coffe_maker', 'television', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'food_snack', 'dishwasher']\nTask:Have snack\nresponse:\n", "output": {"thoughts": "Pour soda into a glass. Get pretzels out of bag. Eat and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTIN] <juice> <freezer>", "[CLOSE] <freezer>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <food_snack>", "[GRAB] <food_snack>", "[CLOSE] <cupboard>", "[DRINK] <cup>", "[EAT] <food_snack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'measuring_cup', 'jelly', 'chair', 'freezer', 'hanger', 'ceiling', 'thread', 'computer', 'mail', 'bookmark', 'deck_of_cards', 'faucet', 'mousepad', 'dining_room', 'broom', 'wall', 'wall_clock', 'box', 'razor', 'doorjamb', 'mouse', 'pencil', 'hairbrush', 'door', 'napkin', 'table', 'light', 'dirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter the Home office, then walk to the phone, pickup the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'toilet', 'bathroom', 'drawing', 'hanger', 'floor', 'bedroom', 'stovefan', 'mat', 'tvstand', 'stereo', 'faucet', 'couch', 'wall', 'nightstand', 'powersocket', 'oven', 'bed', 'photoframe', 'coffe_maker', 'check', 'kettle', 'kitchen_counter', 'dresser', 'home_office', 'light', 'dishwasher', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Put iPhone into speaker holder, open up the Pandora app, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <stereo>", "[PLUGIN] <phone>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'dry_pasta', 'ceilingfan', 'trashcan', 'drawing', 'longboard', 'floor', 'ceiling', 'bedroom', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'shower', 'deck_of_cards', 'couch', 'food_kiwi', 'wall', 'tape', 'basket_for_clothes', 'keys', 'coffe_maker', 'toilet_paper', 'television', 'curtain', 'cd', 'controller', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch. Get comfortable. Switch on the television. Stare at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'television', 'doorjamb', 'towel_rack', 'dresser', 'pajamas', 'sink', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "We sit on the couch facing the tv turn on the television with the remote and watch programs", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'chair', 'food_carrot', 'hanger', 'bathroom_counter', 'floor', 'keyboard', 'mat', 'plate', 'ceilinglamp', 'window', 'mousepad', 'couch', 'food_orange', 'pillow', 'broom', 'wall', 'toy', 'bed', 'toilet_paper', 'television', 'wine', 'curtain', 'food_onion', 'doorjamb', 'dresser', 'spoon', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Sit on sofa. Lie down on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'food_pizza', 'chair', 'walllamp', 'detergent', 'food_carrot', 'drawing', 'floor', 'desk', 'alcohol', 'computer', 'brush', 'mat', 'stovefan', 'toaster', 'tvstand', 'foundation', 'pillow', 'wall', 'toy', 'tape', 'bed', 'keys', 'food_rice', 'curtain', 'pencil', 'comb', 'dresser']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Grab toys off floor.  Put it in a container.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'trashcan', 'dough', 'orchid', 'knife', 'clothes_underwear', 'phone', 'tea_bag', 'envelope', 'mat', 'faucet', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bench', 'powersocket', 'basket_for_clothes', 'food_food', 'food_rice', 'tray', 'curtain', 'after_shave', 'doorjamb', 'pencil', 'laundry_detergent', 'bathroom_cabinet', 'table', 'light', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['scissors', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'newspaper', 'television', 'doorjamb', 'dresser', 'soap', 'instrument_violin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on my couch and watch TV in my living room", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'blender', 'milk', 'orchid', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'thread', 'stovefan', 'mat', 'toaster', 'faucet', 'plate', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'wall_clock', 'oven', 'tray', 'curtain', 'mouse', 'sink', 'napkin', 'bathroom_cabinet', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom, find sink, turn on faucet, put soap in hands, wash and rinse hands,,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <soap>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'fax_machine', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Write report\nresponse:\n", "output": {"thoughts": "Sit at a desk, turn on a computer.  Open an existing document and revise it.  Print a hard copy for mailing.  Email a copy to the client.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[PULL] <mouse>", "[TOUCH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TYPE] <keyboard>", "[TOUCH] <mouse>", "[FIND] <fax_machine>", "[SWITCHON] <fax_machine>", "[SWITCHOFF] <fax_machine>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'chair', 'jelly', 'trashcan', 'toothbrush', 'orchid', 'cpuscreen', 'character', 'bookshelf', 'button', 'mat', 'deck_of_cards', 'faucet', 'shower', 'pillow', 'powersocket', 'bench', 'photoframe', 'bed', 'novel', 'coin', 'kitchen_counter', 'headset', 'chessboard', 'comb', 'dresser', 'table', 'coffee', 'creditcard']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to the entrance hall. Find the telephone. Grab the telephone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'orchid', 'drawing', 'cpuscreen', 'bills', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'mousepad', 'clothes_socks', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'mouse', 'shoes', 'comb', 'door', 'light']\nTask:Check email\nresponse:\n", "output": {"thoughts": "Turn on laptop.  Connect to Internet.  Type in the url.  Check for new email messages.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <mail>", "[GRAB] <mail>", "[READ] <mail>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'knifeblock', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'faucet', 'couch', 'dustpan', 'pillow', 'towel', 'wall', 'nightstand', 'oven', 'box', 'bed', 'keys', 'check', 'toilet_paper', 'television', 'kitchen_counter', 'dresser', 'instrument_guitar', 'soap', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to living room.  Find sofa.  Sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'chair', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'purse', 'brush', 'mat', 'stereo', 'clothes_socks', 'closetdrawer', 'couch', 'mousepad', 'pillow', 'bowl', 'wall', 'powersocket', 'photoframe', 'tray', 'doorjamb', 'pencil', 'shoes', 'cd', 'dresser', 'door', 'bathroom_cabinet', 'light']\nTask:Take off shoes\nresponse:\n", "output": {"thoughts": "Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <shoes> <dresser>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[PUTIN] <shoes> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'coffe_maker', 'television', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'home_office', 'table', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the living room and find the book you want to read. Take it and then go to the couch. Sit on it, open the book and start reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'chair', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'purse', 'computer', 'food_bacon', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'television', 'food_donut', 'wine', 'tray', 'mouse', 'kitchen_counter', 'door', 'hairdryer']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sitting and  reading a hardcover book in a recliner", "actions": ["[RUN] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[RUN] <chair>", "[SIT] <chair>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[DROP] <bookmark>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'freezer', 'walllamp', 'bathroom', 'nail_polish', 'detergent', 'trashcan', 'drawing', 'floor', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'table_cloth', 'stereo', 'mousepad', 'broom', 'wall', 'nightstand', 'powersocket', 'oven', 'television', 'rag', 'mouse', 'headset', 'door', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my bedroom.  I locate my stereo and turn it on.  I put on my headphones and sit to listen.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[WALK] <stereo>', '[SWITCHON] <stereo>', '[FIND] <headset>', '[GRAB] <headset>', '[PUTON] <headset>', '[FIND] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'box', 'cup', 'photoframe', 'television', 'doorjamb', 'coffee_pot', 'dresser', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch. Get comfortable. Switch on the television. Stare at the television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'scissors', 'blender', 'chair', 'walllamp', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'character', 'desk', 'computer', 'food_bacon', 'mat', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'cup', 'photoframe', 'newspaper', 'television', 'ice', 'doorjamb', 'diary', 'console', 'towel_rack']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Drink a glass of water in the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['food_salt', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'stamp', 'bookshelf', 'clothes_jacket', 'mop_bucket', 'toaster', 'mat', 'brush', 'shoe-shine_kit', 'laptop', 'couch', 'pillow', 'wall', 'oven', 'cup', 'food_cheese', 'photoframe', 'check', 'television', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'bathroom_cabinet']\nTask:Eat cheese\nresponse:\n", "output": {"thoughts": "Open fridge.  Get cheese.  Eat it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[EAT] <food_cheese>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'filing_cabinet', 'tablelamp', 'crayon', 'woman', 'wallshelf', 'shampoo', 'floor', 'phone', 'desk', 'man', 'bookshelf', 'closetdrawer', 'couch', 'dining_room', 'nightstand', 'child', 'bed', 'photoframe', 'standingmirror', 'toilet_paper', 'oven_mitts', 'towel_rack', 'kitchen_counter', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'vase']\nTask:Say goodbye to guests leaving\nresponse:\n{\n    \"thoughts\": Tell them bye,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[FIND] <man>', '[TOUCH] <man>', '[FIND] <woman>', '[TOUCH] <woman>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'chair', 'walllamp', 'dough', 'woman', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'man', 'keyboard', 'computer', 'mat', 'tvstand', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'television', 'coin', 'wooden_spoon', 'doorjamb', 'clothes_dress', 'clothes_scarf', 'table', 'coffee']\nTask:Hang with friends\nresponse:\n", "output": {"thoughts": "Sit with friends. Talk", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'freezer', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'table_cloth', 'ceilinglamp', 'faucet', 'couch', 'clothes_hat', 'pillow', 'wall', 'oven', 'bed', 'photoframe', 'television', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I would walk into the living room. Then sit on the couch and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Do homework\nresponse:\n", "output": {"thoughts": "I do my online homework on the computer.  I use Pearson My Math Lab for math class and excel for my engineering class.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'orchid', 'bills', 'cpuscreen', 'knifeblock', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'microphone', 'form', 'towel', 'wall', 'oven', 'food_food', 'television', 'ice', 'cards', 'mouse', 'pencil', 'cupboard', 'door', 'clothes_scarf', 'table']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Wipe down computer desk and monitor,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <desk>', '[FIND] <computer>', '[WIPE] <computer>', '[WIPE] <desk>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'laptop', 'mousepad', 'oil', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Write  school paper\nresponse:\n", "output": {"thoughts": "Writing  and drafting a paper for  class work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'filing_cabinet', 'tablelamp', 'blender', 'mop', 'bathroom', 'detergent', 'toothbrush', 'bills', 'hanger', 'floor', 'bedroom', 'envelope', 'toaster', 'stovefan', 'clothes_socks', 'mousepad', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'toilet_paper', 'towel_rack', 'clothes_dress', 'bag', 'home_office', 'light', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'computer', 'food_bacon', 'mat', 'tvstand', 'stereo', 'foundation', 'mousepad', 'couch', 'pillow', 'wall', 'photoframe', 'check', 'video_game_controller', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'hairdryer', 'bag', 'home_office', 'dishwasher']\nTask:Relax\nresponse:\n", "output": {"thoughts": "Lie down on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'freezer', 'knife', 'drawing', 'needle', 'wallshelf', 'floor', 'phone', 'ceiling', 'dvd_player', 'food_bacon', 'ceilinglamp', 'laptop', 'plate', 'shower', 'tvstand', 'tooth_paste', 'wall', 'tape', 'cup', 'novel', 'food_food', 'television', 'coffee_filter', 'hair', 'cupboard', 'table', 'board_game', 'clothes_shirt']\nTask:Eat\nresponse:\n{\n    \"thoughts\": I get my prepared dinner set on the dining table, sit down on chair, pick up utensils, eat food,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[PUTBACK] <food_food> <table>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <knife>', '[GRAB] <knife>', '[FIND] <plate>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'bathroom', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'purse', 'computer', 'food_bacon', 'toaster', 'mat', 'closetdrawer', 'candle', 'wall', 'nightstand', 'television', 'curtain', 'doorjamb', 'centerpiece', 'cards', 'shoes', 'dresser', 'hairbrush', 'bathroom_cabinet', 'clothes_shirt', 'shoe_rack']\nTask:Put on coat\nresponse:\n", "output": {"thoughts": "Open closet and grab coat.  Put on coat and fasten the front closed.  Close closet.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'chair', 'detergent', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'shower', 'mousepad', 'dustpan', 'dining_room', 'wall', 'powersocket', 'oven', 'check', 'television', 'doorjamb', 'hairbrush', 'controller', 'sink', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk to living room.  Find light switch.  Switch on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'cat', 'food_butter', 'hanger', 'floor', 'ceiling', 'food_cake', 'ceilinglamp', 'deck_of_cards', 'tvstand', 'homework', 'mousepad', 'closetdrawer', 'teeth', 'food_jam', 'nightstand', 'powersocket', 'bed', 'check', 'video_game_controller', 'mouthwash', 'towel_rack', 'mouse', 'console', 'cd', 'cupboard', 'instrument_guitar', 'coffee', 'shoe_rack']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Clean the welcome mat", "actions": ["[WALK] <home_office>", "[WALK] <vacuum_cleaner>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'band-aids', 'bathroom', 'trashcan', 'phone', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'picture', 'window', 'couch', 'bowl', 'broom', 'wall', 'oven', 'bench', 'cup', 'food_food', 'oven_mitts', 'colander', 'curtain', 'pencil', 'kitchen_counter', 'clothes_dress', 'sink', 'dishwasher', 'soap', 'instrument_violin']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find Cat in living room.  Walk to cat.  Pet cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'mop', 'measuring_cup', 'dog', 'walllamp', 'bathroom', 'food_butter', 'woman', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'man', 'table_cloth', 'electric_shaver', 'dining_room', 'child', 'mouthwash', 'razor', 'rag', 'doorjamb', 'lighter', 'towel_rack', 'mouse', 'shoes', 'hair', 'food_steak', 'soap']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet guests when they knock on the door and invite them to enter my home", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink']\nTask:Work\nresponse:\n{\n    \"thoughts\": working on computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <button>', '[TURNTO] <button>', '[LOOKAT] <button>', '[FIND] <mouse>', '[GRAB] <mouse>', '[GRAB] <button>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'walllamp', 'detergent', 'milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'napkin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Find the remote, turn on the TV, watch.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'trashcan', 'drawing', 'floor', 'character', 'toaster', 'plate', 'tvstand', 'foundation', 'dish_soap', 'tooth_paste', 'wall', 'cup', 'coffe_maker', 'food_food', 'food_onion', 'centerpiece', 'cards', 'kettle', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'door', 'sink', 'light', 'dishwasher', 'board_game', 'clothes_shirt', 'creditcard']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Wash a plate using the dishwasher in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sponge', 'fork', 'pot', 'chair', 'detergent', 'floor', 'ceiling', 'dvd_player', 'toaster', 'mat', 'faucet', 'dish_soap', 'food_peanut_butter', 'form', 'bowl', 'dining_room', 'wall', 'nightstand', 'cup', 'bed', 'coffe_maker', 'mouthwash', 'check', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'controller', 'sink', 'clothes_shirt']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Fill sink with warm, sudsy water. Place dirty dishes in soapy water. With a washcloth or sponge scrub dishes clean. Rinse under clean warm water. Place in dish rack to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sink>", "[PUTOBJBACK] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <bowl>", "[GRAB] <bowl>", "[SCRUB] <bowl>", "[RINSE] <bowl>", "[FIND] <dishrack>", "[PUTBACK] <bowl> <dishrack>", "[FIND] <bowl>", "[GRAB] <bowl>", "[SCRUB] <bowl>", "[RINSE] <bowl>", "[PUTBACK] <bowl> <dishrack>", "[RINSE] <sponge>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <sponge>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'walllamp', 'ceilingfan', 'orchid', 'hanger', 'longboard', 'floor', 'ceiling', 'character', 'bedroom', 'folder', 'bookshelf', 'toaster', 'ceilinglamp', 'plate', 'dish_soap', 'candle', 'wall', 'oven', 'bench', 'box', 'cup', 'clothes_gloves', 'coffe_maker', 'tray', 'console', 'maindoor', 'door', 'napkin', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'juice', 'measuring_cup', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'bedroom', 'purse', 'computer', 'mat', 'tvstand', 'shower', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'photoframe', 'television', 'doorjamb', 'shaving_cream', 'mouse', 'home_office']\nTask:Arrange furniture\nresponse:\n", "output": {"thoughts": "Move furniture where you want it.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[PUSH] <coffee_table>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'freezer', 'bathroom', 'food_carrot', 'knifeblock', 'food_egg', 'hanger', 'floor', 'phone', 'stovefan', 'mat', 'plate', 'stereo', 'dish_soap', 'food_noodles', 'closetdrawer', 'mousepad', 'dustpan', 'cup', 'food_food', 'coffe_maker', 'check', 'oven_mitts', 'water', 'food_rice', 'curtain', 'mouse', 'hairbrush', 'sink', 'home_office']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": walk to kitchen, fins soap, arrange dishes, wash dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[FIND] <dish_soap>', '[FIND] <water>', '[GRAB] <plate>', '[GRAB] <dish_soap>', '[WASH] <plate>', '[RINSE] <plate>', '[DROP] <plate>']\n}\n", "input": "\nScene_Objects:['glue', 'sheets', 'filing_cabinet', 'food_bread', 'food_salt', 'microwave', 'chair', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'mail', 'stovefan', 'ceilinglamp', 'mat', 'stereo', 'couch', 'wall', 'powersocket', 'cup', 'check', 'mouthwash', 'food_donut', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'table', 'teeth']\nTask:Play games\nresponse:\n", "output": {"thoughts": "I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.", "actions": ["[WALK] <home_office>", "[WALK] <board_game>", "[FIND] <board_game>", "[GRAB] <board_game>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <board_game> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'chair', 'walllamp', 'blow_dryer', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'ceiling', 'thread', 'ceilinglamp', 'laptop', 'faucet', 'mousepad', 'couch', 'candle', 'food_jam', 'pillow', 'wall', 'nightstand', 'tape', 'television', 'tray', 'after_shave', 'food_onion', 'shoes', 'cupboard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I sit in the sofa and I relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'desk', 'brush', 'mat', 'ceilinglamp', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'wall', 'powersocket', 'cup', 'photoframe', 'coffee_filter', 'food_rice', 'doorjamb', 'towel_rack', 'dresser', 'clothes_dress', 'door', 'light', 'table']\nTask:Change clothes\nresponse:\n", "output": {"thoughts": "Switch your clothes.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTON] <clothes_dress>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'laptop', 'mat', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'cup', 'photoframe', 'television', 'doorjamb', 'mouse', 'door', 'creditcard']\nTask:Play on laptop\nresponse:\n{\n    \"thoughts\": Sit down on the couch. Place laptop on lap. Open laptop. Turn on laptop. Choose selected program. Play selected program. Close game. Close laptop.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <couch>', '[SIT] <couch>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <laptop>', '[PUTOBJBACK] <laptop>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'tablelamp', 'juice', 'microwave', 'freezer', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'powersocket', 'food_vegetable', 'food_food', 'television', 'ice', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I prepare meals and cook them on the stove.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[FIND] <oven>", "[SWITCHON] <oven>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'walllamp', 'trashcan', 'drawing', 'floor', 'ceiling', 'button', 'toaster', 'window', 'mousepad', 'clothes_pants', 'wall', 'basket_for_clothes', 'coffe_maker', 'check', 'television', 'toothbrush_holder', 'ice', 'diary', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'pajamas', 'napkin', 'coffee_cup', 'dishwasher', 'soap', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to the kitchen at the coffee maker, grab coffee, pour coffee into coffee maker, put coffee into coffee maker, push coffee maker's button, wait to complete, grab coffee cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[OPEN] <coffe_maker>', '[POUR] <coffee> <coffe_maker>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'tray', 'food_onion', 'doorjamb', 'shoes', 'dresser', 'sink', 'table', 'tooth_paste']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the living room, I settle on the sofa, I turn on the television, change the channel", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[DROP] <shoes>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'bookshelf', 'brush', 'mat', 'stereo', 'picture', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'shelf', 'cup', 'television', 'wine', 'curtain', 'food_onion', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'home_office', 'table']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to living room. Walk to radio. Switch on the radio. Walk to sofa and sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Using desktop computer in home office", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[WALK] <desk>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'purse', 'stovefan', 'ceilinglamp', 'stereo', 'faucet', 'window', 'balanceball', 'electric_shaver', 'couch', 'wall', 'bench', 'food_food', 'television', 'curtain', 'diary', 'mouse', 'dresser', 'hair', 'bathroom_cabinet', 'table', 'home_office', 'vase', 'teeth', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'clothes_socks', 'couch', 'pillow', 'wall', 'powersocket', 'box', 'cup', 'photoframe', 'television', 'food_rice', 'doorjamb', 'towel_rack', 'kitchen_counter', 'door', 'home_office', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'faucet', 'mousepad', 'wall', 'nightstand', 'cup', 'coffe_maker', 'doorjamb', 'mouse', 'chessboard', 'controller', 'bathroom_cabinet', 'light']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": enter office, find computer, turn it on,,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TOUCH] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'mop', 'toilet', 'freezer', 'walllamp', 'bathroom', 'cat', 'food_butter', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'iron', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'toy', 'tape', 'photoframe', 'television', 'food_rice', 'doorjamb', 'cards', 'door']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'trashcan', 'ceilingfan', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'laptop', 'balanceball', 'mousepad', 'pillow', 'wall', 'cup', 'bed', 'television', 'razor', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'spoon', 'light', 'vase']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": push any ligth button to turn it on, for example in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'tablelamp', 'chair', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'mousepad', 'couch', 'form', 'wall', 'oven', 'powersocket', 'tape', 'cup', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'vacuum_cleaner', 'shoes', 'door', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I go to my home office. I walk towards my desk. I see the phone. I pick it up.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'toilet', 'chair', 'walllamp', 'floor', 'ceiling', 'phone', 'keyboard', 'envelope', 'toaster', 'ceilinglamp', 'stereo', 'faucet', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'clothes_gloves', 'novel', 'check', 'video_game_controller', 'television', 'kitchen_counter', 'hairdryer', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Walk to the sofa and sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'tablelamp', 'microwave', 'freezer', 'food_butter', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'desk', 'keyboard', 'stovefan', 'mat', 'tvstand', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'food_cheese', 'photoframe', 'food_food', 'television', 'wine', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and walk over the cabinent. I pull out a pan and walk to the stove. I put the pan on the stove and walk over to the fridge. I pull out out meat from the fridge and walk back over the stove. Then I put the meat in the pan.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[WALK] <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sheets', 'chair', 'ground_coffee', 'walllamp', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'tvstand', 'stereo', 'window', 'clothes_hat', 'wall', 'nightstand', 'bench', 'box', 'food_cheese', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'hair', 'door', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk to living room.  Find light switch.  Switch on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'walllamp', 'drawing', 'needle', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'character', 'computer', 'toaster', 'mat', 'tvstand', 'deck_of_cards', 'stovefan', 'ceilinglamp', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'television', 'tray', 'doorjamb', 'pencil', 'conditioner', 'door', 'table', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Find couch. Sit on couch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'dvd_player', 'desk', 'bedroom', 'bookshelf', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'tray', 'curtain', 'controller', 'sink', 'bathroom_cabinet', 'table', 'home_office', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dvd_player>', '[FIND] <dvd_player>', '[TURNTO] <dvd_player>', '[LOOKAT] <dvd_player>', '[SWITCHON] <dvd_player>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'laptop', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'soap']\nTask:Work\nresponse:\n", "output": {"thoughts": "Sit in front of the computer, switch on the computer and start typing (Work)", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'ceilingfan', 'toothbrush', 'knifeblock', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'piano_bench', 'toaster', 'faucet', 'window', 'mousepad', 'clothes_pants', 'wall', 'nightstand', 'pasta', 'photoframe', 'toothbrush_holder', 'shoes', 'dresser', 'clothes_dress', 'door', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'tooth_paste', 'laser_pointer']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": go to bathroom. grab toothbrush. put toothpaste on toothbrush. brush teeth. turn on faucet. rinse toothbrush and mouth. put toothbrush away.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'food_butter', 'drawing', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'tvstand', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'photoframe', 'bed', 'coffe_maker', 'television', 'food_donut', 'tray', 'curtain', 'doorjamb', 'pencil', 'dresser', 'napkin', 'dirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'cd_player', 'chair', 'knifeblock', 'music_stand', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'envelope', 'food_bacon', 'mop_bucket', 'stovefan', 'ceilinglamp', 'shower', 'nightstand', 'bed', 'photoframe', 'coffe_maker', 'tray', 'curtain', 'lighter', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to the living and turn to the music stand. Switch on the CD player to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <music_stand>', '[FIND] <music_stand>', '[TURNTO] <music_stand>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'orchid', 'drawing', 'floor', 'ceiling', 'toaster', 'food_noodles', 'mousepad', 'food_peanut_butter', 'wall', 'oven', 'cup', 'bed', 'photoframe', 'coffe_maker', 'toilet_paper', 'television', 'coffee_filter', 'ice', 'curtain', 'console', 'kitchen_counter', 'cupboard', 'sink', 'clothes_scarf', 'dishwasher', 'soap', 'dirt', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[PUTBACK] <coffee> <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'coffee_table', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit down, turn on the TV with the remote control that is on the coffee table.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <coffee_table>', '[TURNTO] <coffee_table>', '[LOOKAT] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'jelly', 'walllamp', 'drawing', 'knife', 'wallshelf', 'orchid', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'candle', 'pillow', 'wall', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'sink']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to Living room. Find Television. Walk to sofa. Sit on sofa. Find Remote. Grab remote. Turn to television. Look at television. Switch on Television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'toilet', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'ceiling', 'cleaning_solution', 'mop_bucket', 'table_cloth', 'faucet', 'shower', 'iron', 'clothes_socks', 'pillow', 'wall', 'food_food', 'mirror', 'wine', 'doorjamb', 'coffee_pot', 'paper_towel', 'cupboard', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'dishwasher', 'scrabble']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[FIND] <cleaning_solution>', '[RINSE] <cleaning_solution>', '[WIPE] <mirror>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'comb', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office, walk to desk, sit in chair, turn on computer, look at computer, click mouse and type on keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'glue', 'jelly', 'freezer', 'detergent', 'food_carrot', 'orchid', 'clothes_underwear', 'hanger', 'floor', 'bedroom', 'thread', 'bookmark', 'mat', 'stereo', 'shower', 'mousepad', 'oil', 'nightstand', 'light_bulb', 'coffe_maker', 'television', 'tray', 'pencil', 'shoes', 'bathroom_cabinet', 'dishwasher', 'colander', 'dirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Go to the living room.  Find the telephone.  Lift up the receiver.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'piano_bench', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'stereo', 'deck_of_cards', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'television', 'food_rice', 'mouse', 'door', 'table', 'laser_pointer']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to the office and sit on the chair, then work on the computer", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'ground_coffee', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'toaster', 'plate', 'faucet', 'foundation', 'mousepad', 'food_jam', 'food_kiwi', 'wall', 'pasta', 'powersocket', 'food_food', 'coffe_maker', 'rag', 'shaving_cream', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Take a dish and put it in the dishwasher to be washed,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'television', 'oven_mitts', 'tray', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <chair>", "[SIT] <chair>", "[TURNTO] <television>", "[WATCH] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'band-aids', 'drawing', 'orchid', 'kitchen_cabinet', 'floor', 'bathroom_counter', 'phone', 'bookshelf', 'computer', 'stovefan', 'mat', 'plate', 'ceilinglamp', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'bench', 'powersocket', 'oven', 'cutting_board', 'curtain', 'mouse', 'table', 'dishwasher', 'light']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the kitchen cabinet. Grab the plate and put it on the table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['juice', 'walllamp', 'nail_polish', 'detergent', 'dough', 'drawing', 'food_egg', 'tea_bag', 'bookshelf', 'remote_control', 'alcohol', 'button', 'mop_bucket', 'table_cloth', 'stovefan', 'tvstand', 'shoe-shine_kit', 'shower', 'stereo', 'deck_of_cards', 'mousepad', 'powersocket', 'check', 'rag', 'towel_rack', 'pencil', 'conditioner', 'door', 'bathroom_cabinet', 'soap']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'walllamp', 'drawing', 'floor', 'desk', 'tea_bag', 'computer', 'clothes_jacket', 'toaster', 'mat', 'mousepad', 'closetdrawer', 'towel', 'tape', 'basket_for_clothes', 'bed', 'curtain', 'shaving_cream', 'towel_rack', 'mouse', 'kitchen_counter', 'chessboard', 'dresser', 'hair', 'laundry_detergent', 'bathroom_cabinet', 'light', 'board_game', 'dirt']\nTask:Open front door\nresponse:\n{\n    \"thoughts\": Walk over to the front door, turn the knob, and open the door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'food_noodles', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'headset', 'dresser', 'food_steak', 'sink']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "change the tv channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'chair', 'blow_dryer', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'toaster', 'laptop', 'deck_of_cards', 'faucet', 'food_noodles', 'closetdrawer', 'food_orange', 'wall', 'nightstand', 'clothes_gloves', 'coffe_maker', 'toilet_paper', 'food_food', 'curtain', 'towel_rack', 'console', 'kitchen_counter', 'shoes', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Making coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['glue', 'food_pizza', 'toilet', 'chair', 'bathroom_counter', 'floor', 'ceiling', 'cleaning_solution', 'food_bacon', 'mop_bucket', 'toaster', 'stovefan', 'faucet', 'shower', 'mousepad', 'dustpan', 'pillow', 'wall', 'nightstand', 'powersocket', 'coffe_maker', 'console', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'pajamas', 'sink', 'napkin', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Washing hands", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <cleaning_solution>", "[SQUEEZE] <cleaning_solution>", "[SCRUB] <hands_both>", "[SWITCHON] <faucet>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'scissors', 'freezer', 'ceilingfan', 'knifeblock', 'knife', 'hanger', 'phone', 'desk', 'clothes_jacket', 'toaster', 'ceilinglamp', 'electric_shaver', 'window', 'mousepad', 'couch', 'candle', 'cutting_board', 'food_cheese', 'food_rice', 'food_onion', 'pencil', 'cd', 'comb', 'door', 'maindoor', 'napkin', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": You walk to the living room and find the cat. You look at the cat and touch it to pet it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'bathroom', 'cat', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'bedroom', 'food_bacon', 'piano_bench', 'bookmark', 'stovefan', 'ceilinglamp', 'laptop', 'shower', 'stereo', 'mat', 'couch', 'nightstand', 'wall', 'photoframe', 'coffee_filter', 'towel_rack', 'dishrack', 'controller', 'table', 'clothes_skirt', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into room, grab phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'scissors', 'microwave', 'bathroom', 'drawing', 'knifeblock', 'floor', 'bedroom', 'desk', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'closetdrawer', 'wall', 'powersocket', 'newspaper', 'photoframe', 'bed', 'check', 'food_rice', 'ice', 'curtain', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'instrument_guitar', 'board_game']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": go into the kitchen, walk to the sink. Turn on the water and rinse of a plate and clean them off fully with a sponge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <plate>', '[RINSE] <plate>', '[FIND] <sponge>', '[GRAB] <sponge>', '[WASH] <plate>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'food_salt', 'measuring_cup', 'mop', 'chair', 'walllamp', 'cat', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'computer', 'mat', 'clothes_pants', 'oven', 'check', 'toilet_paper', 'oven_mitts', 'food_donut', 'after_shave', 'lighter', 'cards', 'headset', 'instrument_violin', 'instrument_guitar', 'light', 'board_game', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go to the living room then go to the cat and touch it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'walllamp', 'food_butter', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'shower', 'picture', 'mousepad', 'pillow', 'wall', 'oven', 'powersocket', 'bed', 'photoframe', 'food_donut', 'tray', 'mouse', 'pencil', 'bag', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": go to living room,find the text book on the self ,sat on the chair ,open and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'mat', 'ceilinglamp', 'shower', 'clothes_socks', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'oven', 'food_food', 'television', 'tray', 'wooden_spoon', 'food_onion', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'hairdryer', 'door', 'dishwasher', 'clothes_skirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, put the groceries in the fridge and close it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'bathroom', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'mousepad', 'dining_room', 'wall', 'light_bulb', 'check', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'clothes_dress', 'door', 'table', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'freezer', 'band-aids', 'walllamp', 'nail_polish', 'orchid', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'keyboard', 'toaster', 'ceilinglamp', 'plate', 'stereo', 'mat', 'iron', 'food_peanut_butter', 'food_jam', 'wall', 'oven', 'check', 'food_food', 'food_rice', 'ice', 'food_onion', 'console', 'kitchen_counter', 'table']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the kitchen cabinet. Grab the plate and put it on the table.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'toilet', 'dough', 'toothbrush', 'drawing', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'stereo', 'shower', 'window', 'oil', 'couch', 'bowl', 'wall', 'oven', 'toy', 'cup', 'bed', 'food_food', 'television', 'wooden_spoon', 'towel_rack', 'comb', 'cupboard', 'table', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Find couch. Sit on couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'band-aids', 'detergent', 'food_carrot', 'milk', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'button', 'piano_bench', 'toaster', 'plate', 'faucet', 'drying_rack', 'shower', 'dish_soap', 'closetdrawer', 'clothes_pants', 'wall', 'oven', 'food_cheese', 'newspaper', 'food_food', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'shoe_rack']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I walk in to the Kitchen. I find a plate, a sponge, and the dish soap. I find the faucet. I turn the faucet on. I pour water on the sponge and plate. I add dish soap to the sponge. I put the dish soap back. I scrub the plate. I rinse the plate. I place the plate on the drying rack. I rinse the sponge. I turn the faucet off.", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <sponge>", "[GRAB] <sponge>", "[RINSE] <sponge>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>", "[RINSE] <plate>", "[FIND] <drying_rack>", "[PUTBACK] <plate> <plate>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'scissors', 'juice', 'food_salt', 'microwave', 'chair', 'food_butter', 'toothbrush', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'button', 'brush', 'ceilinglamp', 'plate', 'wall', 'tape', 'food_food', 'check', 'food_donut', 'wooden_spoon', 'rag', 'door', 'clothes_scarf', 'napkin', 'table', 'light', 'dishwasher']\nTask:Eat\nresponse:\n{\n    \"thoughts\": Serve food on dinning table in appropriate utensils, sit on the chair, eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <plate>', '[PUTBACK] <food_food> <plate>', '[GRAB] <food_food>', '[PUTOBJBACK] <food_food>', '[FIND] <fork>', '[GRAB] <fork>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'ground_coffee', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Sit down at desk in home office, turn on computer, open browser on computer, browse internet", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'knifeblock', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'television', 'curtain', 'doorjamb', 'cd', 'pajamas', 'door', 'table', 'light', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sitting on the couch watching my favorite television show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'beer', 'freezer', 'walllamp', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'keyboard', 'toaster', 'mat', 'tvstand', 'couch', 'food_orange', 'pillow', 'wall', 'basket_for_clothes', 'food_cheese', 'food_food', 'television', 'after_shave', 'towel_rack', 'pencil', 'kitchen_counter', 'cupboard', 'controller', 'bathroom_cabinet', 'table', 'dirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I leave the bags with groceries near the fridge, open the door and put the groceries in the shelves.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mail', 'stovefan', 'ceilinglamp', 'mat', 'mousepad', 'dustpan', 'wall', 'bed', 'coffe_maker', 'lighter', 'coffee_pot', 'mouse', 'hair', 'door', 'napkin', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Go find a computer, turn it on, click google, and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['water_glass', 'scissors', 'blender', 'chair', 'walllamp', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'character', 'desk', 'computer', 'food_bacon', 'mat', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'cup', 'photoframe', 'newspaper', 'television', 'ice', 'doorjamb', 'diary', 'console', 'towel_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Drink a glass of water in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'ceilinglamp', 'shoe-shine_kit', 'faucet', 'clothes_pants', 'dining_room', 'wall', 'tape', 'basket_for_clothes', 'bed', 'coffe_maker', 'chessboard', 'clothes_dress', 'door', 'home_office', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'toilet', 'chair', 'band-aids', 'orchid', 'drawing', 'kitchen_cabinet', 'floor', 'table_cloth', 'ceilinglamp', 'plate', 'mat', 'tvstand', 'electric_shaver', 'candle', 'microphone', 'wall', 'cup', 'photoframe', 'food_food', 'food_rice', 'tray', 'food_onion', 'kitchen_counter', 'chessboard', 'cupboard', 'home_office', 'table', 'laser_pointer']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Grab plates and forks. Walk into dining room. Put plates and forks on table.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <fork> <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'food_salt', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'powersocket', 'toy', 'tape', 'light_bulb', 'check', 'food_food', 'television', 'curtain', 'doorjamb', 'mouse', 'pencil', 'dresser', 'conditioner', 'spoon', 'light', 'table', 'colander']\nTask:Re arrange office\nresponse:\n{\n    \"thoughts\": Walk in.Turn on light.Move furniture. Change the room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[PUSH] <desk>', '[WALK] <chair>', '[PUSH] <chair>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'freezer', 'milk', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'alcohol', 'computer', 'toaster', 'shoe-shine_kit', 'faucet', 'pillow', 'dining_room', 'wall', 'cup', 'bed', 'check', 'toilet_paper', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'clothes_dress', 'food_steak', 'sink', 'bathroom_cabinet']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk into the kitchen. I reach into the cabinet and pick up a water glass. Turn on the faucet and place the glass under the water stream. Bring the glass of water to lips and drink.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'couch', 'wall', 'nightstand', 'oven', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'food_onion', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'electrical_outlet']\nTask:Toast bread\nresponse:\n{\n    \"thoughts\": Place two slices of bread in the toaster. Select medium toast on toaster.  Push down button to begin toasting.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[CLOSE] <freezer>', '[WALK] <electrical_outlet>', '[FIND] <toaster>', '[PLUGIN] <toaster>', '[PUTBACK] <food_bread> <toaster>', '[SWITCHON] <electrical_outlet>', '[TURNTO] <toaster>', '[LOOKAT] <toaster>', '[SWITCHON] <toaster>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'bathroom', 'detergent', 'bathroom_counter', 'button', 'toaster', 'brush', 'stereo', 'shower', 'faucet', 'food_orange', 'pillow', 'oven', 'photoframe', 'check', 'television', 'tray', 'kettle', 'mouse', 'cupboard', 'conditioner', 'door', 'home_office', 'table', 'dishwasher', 'clothes_skirt', 'soap']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room, look for radio, grab on button", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'blender', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'keyboard', 'mat', 'laptop', 'tvstand', 'picture', 'window', 'clothes_socks', 'couch', 'pillow', 'food_kiwi', 'wall', 'wall_clock', 'cutting_board', 'bed', 'television', 'toothbrush_holder', 'curtain', 'food_onion', 'doorjamb', 'door', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room. I cross the room to the couch. I sit down and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'wooden_spoon', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to my home office. I sit on the chair right in front of my computer. I turn on the computer. Use the mouse to open the web browser. Use the keyboard to type the URL that I want to visit.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'food_butter', 'drawing', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'tvstand', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'photoframe', 'bed', 'coffe_maker', 'television', 'food_donut', 'tray', 'curtain', 'doorjamb', 'pencil', 'dresser', 'napkin', 'dirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'tea_bag', 'bookshelf', 'mat', 'deck_of_cards', 'clothes_socks', 'closetdrawer', 'couch', 'mousepad', 'wall', 'nightstand', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Coming home turning on lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'orchid', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'microphone', 'pillow', 'wall', 'wall_clock', 'photoframe', 'television', 'toothbrush_holder', 'food_donut', 'doorjamb', 'shaving_cream', 'mouse', 'door', 'sink', 'clothes_scarf', 'napkin', 'table']\nTask:Straighten pictures on wall\nresponse:\n{\n    \"thoughts\": Move the pictures/paintings until they are straight.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PUSH] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PUSH] <drawing>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'juice', 'toilet', 'chair', 'nail_polish', 'needle', 'clothes_underwear', 'phone', 'keyboard', 'bookmark', 'stovefan', 'ceilinglamp', 'toaster', 'shower', 'foundation', 'clothes_hat', 'pillow', 'oven', 'powersocket', 'tape', 'bed', 'coffe_maker', 'television', 'curtain', 'towel_rack', 'mouse', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the radio. Turn on the radio. Listen to the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'food_carrot', 'drawing', 'hanger', 'phone', 'ceiling', 'desk', 'computer', 'toaster', 'mat', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'towel', 'cup', 'food_cheese', 'keys', 'tray', 'after_shave', 'towel_rack', 'mouse', 'cupboard', 'sink', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'tvstand', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'check', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'clothes_dress', 'door', 'conditioner', 'sink', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "My cat looks like it wants attention, so I will go pet it on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'pot', 'toilet', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'kettle', 'shaving_cream', 'kitchen_counter', 'chessboard', 'dresser', 'controller', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I get home from work and walk into the living room. I see the couch and go to sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'bathroom', 'orchid', 'food_egg', 'cpuscreen', 'floor', 'phone', 'bedroom', 'desk', 'stovefan', 'toaster', 'ceilinglamp', 'shower', 'mat', 'food_peanut_butter', 'wall', 'pasta', 'powersocket', 'oven', 'cup', 'bed', 'after_shave', 'food_onion', 'chessboard', 'dresser', 'cupboard', 'clothes_scarf', 'home_office']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my living room, find my cat, scrub it using my hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'dog', 'chair', 'bathroom', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'mat', 'ceilinglamp', 'shower', 'candle', 'wall', 'bed', 'clothes_gloves', 'food_food', 'water', 'food_rice', 'wooden_spoon', 'curtain', 'console', 'dishrack', 'pajamas', 'light', 'clothes_skirt', 'soap', 'dirt']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'microwave', 'bathroom', 'knife', 'floor', 'ceiling', 'bedroom', 'toaster', 'plate', 'shower', 'dish_soap', 'closetdrawer', 'dustpan', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'check', 'drinking_glass', 'curtain', 'centerpiece', 'shaving_cream', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'dirt', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and turn to the dishwasher, put the knife, glass, fork and plate into the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[TURNTO] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'bathroom', 'trashcan', 'orchid', 'drawing', 'wallshelf', 'floor', 'character', 'bedroom', 'tea_bag', 'keyboard', 'toaster', 'mat', 'stovefan', 'tvstand', 'couch', 'dining_room', 'wall', 'nightstand', 'child', 'bed', 'photoframe', 'novel', 'food_onion', 'after_shave', 'cupboard', 'bathtub', 'napkin', 'bathroom_cabinet', 'board_game']\nTask:Story reading time\nresponse:\n{\n    \"thoughts\": Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[TURNTO] <child>', '[LOOKAT] <child>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I go to my home office. I find the computer along with mechanical keyboard on the desk. I sit on chair. I switch on the computer. I type on keyboard and work on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'shower', 'window', 'clothes_hat', 'dining_room', 'wall', 'bench', 'powersocket', 'photoframe', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'table', 'light', 'tooth_paste']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Pick up phone,\n    \"actions\": ['[WALK] <table>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'microwave', 'band-aids', 'detergent', 'cat', 'drawing', 'orchid', 'ceiling', 'bedroom', 'keyboard', 'thread', 'clothes_jacket', 'toaster', 'iron', 'shower', 'mousepad', 'food_orange', 'broom', 'oven', 'coffee_filter', 'curtain', 'doorjamb', 'centerpiece', 'towel_rack', 'shoes', 'controller', 'door', 'bathroom_cabinet', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat; pet the cat by touching it", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_ice_cream', 'walllamp', 'nail_polish', 'cat', 'orchid', 'drawing', 'floor', 'ceiling', 'purse', 'alcohol', 'mail', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'towel', 'wall', 'oven', 'cup', 'food_cheese', 'television', 'curtain', 'doorjamb', 'door', 'table', 'clothes_shirt']\nTask:Straighten paintings on wall\nresponse:\n{\n    \"thoughts\": Go up to the painting and tilt it which ever way it is uneven until it looks straight.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <wall>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'dough', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'wall', 'photoframe', 'novel', 'television', 'curtain', 'doorjamb', 'towel_rack', 'shoes', 'cupboard', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to the couch in the living room, sit on ti and read the book", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bedroom', 'bookshelf', 'keyboard', 'ceilinglamp', 'faucet', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'food_food', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathtub', 'light', 'table', 'instrument_violin']\nTask:Straighten pictures on wall\nresponse:\n{\n    \"thoughts\": Move the pictures/paintings until they are straight.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PUSH] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PUSH] <drawing>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'wallshelf', 'cpuscreen', 'floor', 'shampoo', 'mop_bucket', 'toaster', 'mat', 'plate', 'shower', 'dish_soap', 'closetdrawer', 'dining_room', 'wall', 'photoframe', 'coffe_maker', 'colander', 'pencil', 'kitchen_counter', 'cupboard', 'clothes_dress', 'food_steak', 'home_office', 'dishwasher', 'soap', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'beer', 'measuring_cup', 'toilet', 'drawing', 'knife', 'hanger', 'floor', 'desk', 'tea_bag', 'keyboard', 'piano_bench', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'foundation', 'clothes_pants', 'towel', 'wall', 'nightstand', 'novel', 'television', 'tray', 'curtain', 'bag', 'table', 'vase', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": BEFORE EATING,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <towel>', '[WIPE] <towel>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'computer', 'mail', 'table_cloth', 'mat', 'tvstand', 'plate', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'curtain', 'after_shave', 'doorjamb', 'cards', 'mouse', 'shoes', 'sink']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I sit in the sofa and I relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'board_game', 'chair', 'ground_coffee', 'detergent', 'knifeblock', 'floor', 'ceiling', 'alcohol', 'toaster', 'faucet', 'picture', 'closetdrawer', 'couch', 'dining_room', 'broom', 'wall', 'coffe_maker', 'check', 'water', 'drinking_glass', 'after_shave', 'rag', 'kitchen_counter', 'shoes', 'cupboard', 'hair', 'sink', 'table', 'tooth_paste']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Walk to faucet. Switch on faucet. Pour water into drinking glass. Switch off faucet. Drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <sink>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[SWITCHOFF] <faucet>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'chair', 'walllamp', 'toothbrush', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'oil', 'closetdrawer', 'couch', 'food_jam', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'check', 'television', 'doorjamb', 'dresser', 'door']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to living room. Walk to and sit at the sofa. Grab remote control and switch on TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'tea', 'microwave', 'chair', 'walllamp', 'blow_dryer', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'plate', 'mousepad', 'food_noodles', 'couch', 'closetdrawer', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'razor', 'television', 'shaving_cream', 'kitchen_counter', 'bathroom_cabinet', 'dishwasher', 'soap', 'teeth']\nTask:Text friends while sitting on couch\nresponse:\n{\n    \"thoughts\": Text while sitting,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[TOUCH] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I walk into my home office and walk up to my desk. I pull out the chair and sit down. I turn on the computer. After the computer is turned on I click on the google icon. I enter Disney Cruise Line into the search bar and browse the internet.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[TURNTO] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TURNTO] <keyboard>", "[LOOKAT] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'purse', 'computer', 'ceilinglamp', 'mat', 'balanceball', 'electric_shaver', 'window', 'mousepad', 'dustpan', 'pillow', 'wall', 'nightstand', 'photoframe', 'food_food', 'doorjamb', 'mouse', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": turn on the computer and browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'filing_cabinet', 'chair', 'band-aids', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'cup', 'photoframe', 'television', 'doorjamb', 'shoes', 'dresser', 'soap']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'glue', 'food_ice_cream', 'food_salt', 'tea', 'drawing', 'knife', 'cpuscreen', 'phone', 'desk', 'computer', 'button', 'plate', 'iron', 'dish_soap', 'mousepad', 'nightstand', 'powersocket', 'curtain', 'kitchen_counter', 'dishrack', 'dresser', 'spoon', 'napkin', 'table', 'dishwasher', 'home_office', 'vase', 'bathroom_cabinet', 'coffee']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <dishrack>', '[PUTBACK] <fork> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishrack>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishrack>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_salt', 'toilet', 'walllamp', 'bathroom', 'food_carrot', 'dough', 'milk', 'phone', 'ceiling', 'bedroom', 'desk', 'keyboard', 'table_cloth', 'picture', 'oil', 'candle', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'oven', 'coffe_maker', 'coin', 'towel_rack', 'pencil', 'dresser', 'clothes_dress', 'clothes_scarf', 'instrument_guitar']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the entrance hall. Find the telephone. Grab the telephone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'television', 'tray', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch youtube\nresponse:\n{\n    \"thoughts\": Go to youtube.Com watch videos,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "I need to complete work on my computer, so I will be pulling out my chair, turning on my computer and begin my task.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'walllamp', 'detergent', 'milk', 'hanger', 'phone', 'toaster', 'tvstand', 'balanceball', 'electric_shaver', 'food_noodles', 'garbage_can', 'mousepad', 'clothes_hat', 'pillow', 'broom', 'food_kiwi', 'bench', 'food_vegetable', 'food_cheese', 'food_food', 'ice', 'wine', 'rag', 'shaving_cream', 'dresser', 'cupboard', 'napkin', 'pantry']\nTask:Organize pantry\nresponse:\n{\n    \"thoughts\": Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <pantry>', '[FIND] <food_food>', '[GRAB] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[FIND] <garbage_can>', '[PUTBACK] <food_food> <garbage_can>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <dry_pasta>', '[GRAB] <dry_pasta>', '[TURNTO] <dry_pasta>', '[LOOKAT] <dry_pasta>', '[PUTOBJBACK] <dry_pasta>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[PUSH] <food_food>', '[FIND] <food_noodles>', '[GRAB] <food_noodles>', '[TURNTO] <food_noodles>', '[LOOKAT] <food_noodles>', '[PUTOBJBACK] <food_noodles>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'stamp', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'dining_room', 'wall', 'toy', 'bed', 'food_food', 'toilet_paper', 'mouthwash', 'doorjamb', 'pencil', 'kitchen_counter', 'comb', 'dresser', 'laundry_detergent', 'napkin', 'home_office', 'bathroom_cabinet', 'dishwasher', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "\"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\"", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[TURNTO] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'mat', 'deck_of_cards', 'shower', 'tvstand', 'food_jam', 'dining_room', 'wall', 'nightstand', 'powersocket', 'food_cheese', 'photoframe', 'novel', 'food_food', 'toilet_paper', 'coffe_maker', 'food_donut', 'doorjamb', 'kitchen_counter', 'dresser', 'door', 'light', 'tooth_paste', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": in my home i walk to the kitchen with some groceries and locate the fridge,then open the fridge put the groceries into the fridge and closed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'tablelamp', 'food_salt', 'measuring_cup', 'walllamp', 'bathroom', 'milk', 'orchid', 'cpuscreen', 'stamp', 'keyboard', 'food_noodles', 'closetdrawer', 'mousepad', 'clothes_pants', 'pasta', 'powersocket', 'food_food', 'video_game_controller', 'check', 'water', 'chessboard', 'dresser', 'cupboard', 'clothes_scarf', 'home_office', 'bathroom_cabinet', 'soap', 'scrabble']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "put soap on hands, scrub it and rinse with water", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'detergent', 'cat', 'drawing', 'floor', 'stamp', 'computer', 'ceilinglamp', 'mat', 'balanceball', 'picture', 'window', 'foundation', 'faucet', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'mouthwash', 'television', 'ice', 'centerpiece', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'table', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_salt', 'chair', 'freezer', 'drawing', 'wallshelf', 'hanger', 'phone', 'folder', 'bedroom', 'desk', 'keyboard', 'purse', 'couch', 'food_peanut_butter', 'clothes_pants', 'dining_room', 'wall', 'water', 'after_shave', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'conditioner', 'sink', 'napkin', 'home_office', 'table', 'tooth_paste']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drinking a glass of water", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'pot', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'oil', 'mousepad', 'food_peanut_butter', 'wall', 'oven', 'powersocket', 'food_food', 'water', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'keyboard', 'purse', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'food_noodles', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "FIRST YOU NEED TO SWITVH ON YOU COMPUTER AND CHECK THE INTERNET CONNECTION.FOR SENDING EMAIL EMAIL ID FOR REQUIRED FOR BOTH SEND AND RECEIVE .SO CLEALY MENTION THE ID ON THE TO PLACE THEN ONLY IT WILL GO CORRECTLY", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'food_egg', 'clothes_underwear', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'bench', 'box', 'television', 'curtain', 'towel_rack', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Changing channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[DROP] <remote_control>']\n}\n", "input": "\nScene_Objects:['glue', 'microwave', 'toilet', 'wallshelf', 'cpuscreen', 'hanger', 'bookshelf', 'computer', 'bookmark', 'mat', 'tvstand', 'shower', 'faucet', 'foundation', 'oil', 'closetdrawer', 'towel', 'powersocket', 'bed', 'novel', 'coin', 'after_shave', 'pencil', 'dresser', 'cupboard', 'clothes_dress', 'napkin', 'table', 'laser_pointer', 'creditcard']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room , find phone ,pick up phone 3", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'blender', 'food_ice_cream', 'food_butter', 'phone', 'character', 'bookshelf', 'brush', 'ceilinglamp', 'plate', 'shower', 'iron', 'stereo', 'toaster', 'oil', 'food_orange', 'pillow', 'cup', 'food_food', 'check', 'oven_mitts', 'food_donut', 'food_onion', 'after_shave', 'mouse', 'hairdryer', 'bag', 'bathroom_cabinet', 'home_office']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <home_office>', '[DROP] <plate>', '[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <home_office>', '[DROP] <cup>', '[WALK] <dining_room>', '[WALK] <fork>', '[FIND] <fork>', '[GRAB] <fork>', '[WALK] <home_office>', '[DROP] <fork>']\n}\n", "input": "\nScene_Objects:['sponge', 'glue', 'tablelamp', 'crayon', 'freezer', 'cat', 'milk', 'wallshelf', 'bathroom_counter', 'bedroom', 'desk', 'bookshelf', 'thread', 'stovefan', 'stereo', 'faucet', 'pillow', 'pasta', 'oven', 'powersocket', 'cup', 'bed', 'coffe_maker', 'after_shave', 'mouse', 'dresser', 'door', 'hairdryer', 'dishwasher', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting a cat may sound simple.Letting the Cat Come to You.Extend a hand or finger and allow the cat a chance to touch her nose to you. When a cat bumps her head into your hand, it's a signal she wants attention.Focusing on Areas with Scent Glands. Feed with food.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <milk>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[TURNTO] <cat>", "[POINTAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'dough', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'clothes_jacket', 'homework', 'mousepad', 'towel', 'wall', 'nightstand', 'bed', 'novel', 'coffe_maker', 'toothbrush_holder', 'ice', 'diary', 'kettle', 'mouse', 'door', 'spoon', 'instrument_violin']\nTask:Study\nresponse:\n{\n    \"thoughts\": Do homework or read books while sitting on the chair.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[PUTOBJBACK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_bread', 'filing_cabinet', 'tablelamp', 'blender', 'crayon', 'measuring_cup', 'microwave', 'freezer', 'bathroom', 'milk', 'needle', 'cpuscreen', 'wallshelf', 'hanger', 'phone', 'computer', 'toaster', 'ceilinglamp', 'newspaper', 'food_food', 'coffe_maker', 'ice', 'curtain', 'after_shave', 'dishrack', 'hairbrush', 'spoon', 'table', 'colander']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk into the living room and walk over to the table with the phone. I pick up the phone and put it to my ear.", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'blender', 'freezer', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'plate', 'iron', 'wall', 'nightstand', 'bed', 'food_food', 'check', 'drinking_glass', 'curtain', 'centerpiece', 'coffee_pot', 'cupboard', 'laundry_detergent', 'door', 'dishwasher', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Robot Drinks the water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>', '[PUTOBJBACK] <drinking_glass>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'water_glass', 'microwave', 'toilet', 'drawing', 'wallshelf', 'ceiling', 'desk', 'bookmark', 'toaster', 'stovefan', 'mat', 'shower', 'couch', 'wall', 'oven', 'coffe_maker', 'coin', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'cupboard', 'controller', 'hairdryer', 'spoon', 'dishwasher', 'creditcard']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen, take a glass from the cupboard, fill with water and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>", "[RINSE] <water_glass>", "[PUTBACK] <water_glass> <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Compute\nresponse:\n{\n    \"thoughts\": Use computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'tablelamp', 'freezer', 'bathroom', 'milk', 'drawing', 'floor', 'ceiling', 'bedroom', 'alcohol', 'computer', 'table_cloth', 'stovefan', 'mat', 'shower', 'oil', 'couch', 'pillow', 'wall', 'cutting_board', 'newspaper', 'food_food', 'television', 'kettle', 'kitchen_counter', 'cupboard', 'clothes_dress', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find groceries. Grab groceries. Find fridge. Walk to fridge. Put groceries in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'measuring_cup', 'walllamp', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'food_noodles', 'couch', 'broom', 'wall', 'newspaper', 'light_bulb', 'novel', 'television', 'curtain', 'doorjamb', 'lighter', 'cupboard', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'filing_cabinet', 'food_bread', 'tablelamp', 'mop', 'toilet', 'freezer', 'walllamp', 'desk', 'keyboard', 'toaster', 'ceilinglamp', 'laptop', 'mousepad', 'dining_room', 'nightstand', 'cup', 'photoframe', 'clothes_gloves', 'novel', 'coffe_maker', 'curtain', 'dresser', 'controller', 'clothes_dress', 'door', 'sink', 'light', 'laser_pointer']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find a cat in the kitchen and touch it", "actions": ["[WALK] <dining_room>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'keyboard', 'computer', 'toaster', 'tvstand', 'floor_lamp', 'window', 'closetdrawer', 'couch', 'towel', 'wall', 'wall_clock', 'food_food', 'doorjamb', 'mouse', 'headset', 'shoes', 'dresser', 'controller', 'bathtub', 'sink', 'home_office', 'light']\nTask:Turn night light on\nresponse:\n{\n    \"thoughts\": I walk to the dresser and in the middle Io to small battery operated lamp and switch it on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <floor_lamp>', '[TURNTO] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'tvstand', 'brush', 'stovefan', 'couch', 'candle', 'pillow', 'wall', 'powersocket', 'toy', 'cup', 'photoframe', 'television', 'food_donut', 'tray', 'doorjamb', 'console', 'pencil', 'kitchen_counter', 'cupboard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Walk to the sofa and sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'alcohol', 'stovefan', 'mat', 'shoe-shine_kit', 'tvstand', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'wall', 'powersocket', 'photoframe', 'food_food', 'check', 'television', 'coffe_maker', 'ice', 'kitchen_counter', 'cupboard', 'hair', 'table', 'scrabble']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'stamp', 'computer', 'brush', 'mat', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'novel', 'coffe_maker', 'television', 'doorjamb', 'vacuum_cleaner', 'hairbrush', 'cupboard', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I pick up a book off the self that I want to read. Then, I sit down on the couch, open the book, and begin reading it.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'chair', 'cat', 'drawing', 'floor', 'tea_bag', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'stovefan', 'window', 'couch', 'food_orange', 'wall', 'food_food', 'television', 'oven_mitts', 'food_onion', 'curtain', 'after_shave', 'comb', 'napkin', 'home_office', 'table', 'dishwasher', 'bathroom_cabinet', 'laser_pointer', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room,find the sofa,sat down in the sofa,take some relaxation.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'toy', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'pencil', 'dresser', 'sink', 'board_game']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room, walk to TV, grab TV remote, turn on TV, put down the remote, walk to sofa, sit in soft, watch TV", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'microwave', 'toilet', 'freezer', 'ceiling', 'clothes_jacket', 'table_cloth', 'toaster', 'stovefan', 'plate', 'faucet', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'table']\nTask:Get out dish\nresponse:\n{\n    \"thoughts\": Open cupboard.  Get out the dish.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[OPEN] <cupboard>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <cupboard>']\n}\n", "input": "\nScene_Objects:['glue', 'toilet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'mail', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cutting_board', 'photoframe', 'bed', 'television', 'doorjamb', 'dishrack', 'dresser', 'door', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to the living room to watch TV", "actions": ["[RUN] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[TOUCH] <remote_control>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[DROP] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'cat', 'ceilingfan', 'food_butter', 'drawing', 'longboard', 'floor', 'ceiling', 'shampoo', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'couch', 'clothes_pants', 'wall', 'wall_clock', 'bench', 'television', 'curtain', 'kettle', 'towel_rack', 'hair', 'pajamas', 'maindoor', 'home_office', 'table', 'vase']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Changing channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[DROP] <remote_control>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'tea', 'drawing', 'wallshelf', 'hanger', 'ceiling', 'tea_bag', 'bookshelf', 'stovefan', 'toaster', 'mat', 'faucet', 'floor_lamp', 'mousepad', 'nightstand', 'box', 'bed', 'photoframe', 'novel', 'food_food', 'coffee_filter', 'food_donut', 'wine', 'after_shave', 'curtain', 'doorjamb', 'clothes_dress', 'conditioner', 'sink']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Turn a light on and read a book in the bedroom.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'bathroom', 'detergent', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'video_game_console', 'keyboard', 'stovefan', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall_clock', 'oven', 'bed', 'coffe_maker', 'video_game_controller', 'television', 'doorjamb', 'cd', 'cupboard', 'door', 'bathtub', 'bathroom_cabinet', 'light', 'home_office', 'table']\nTask:Playing video game\nresponse:\n{\n    \"thoughts\": Playing  video games  on various video game  systems such as wii u  and  xbox,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <video_game_console>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[PUTOBJBACK] <video_game_controller>', '[SWITCHOFF] <video_game_console>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'walllamp', 'milk', 'clothes_underwear', 'character', 'bedroom', 'keyboard', 'clothes_socks', 'candle', 'clothes_pants', 'wall', 'nightstand', 'powersocket', 'basket_for_clothes', 'photoframe', 'bed', 'food_food', 'curtain', 'diary', 'shoes', 'headset', 'laundry_detergent', 'sink', 'home_office', 'table', 'colander', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[DROP] <basket_for_clothes>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[GRAB] <laundry_detergent>", "[PUTOBJBACK] <laundry_detergent>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sponge', 'food_bread', 'blender', 'chair', 'walllamp', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'tvstand', 'window', 'mousepad', 'clothes_socks', 'wall', 'oven', 'cup', 'television', 'mouse', 'cupboard', 'door', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": go to the office, turn computer on, open an browser or any application to open any website.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'tape', 'photoframe', 'television', 'doorjamb', 'kettle', 'mouse', 'dresser']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room, find the television, switch on the television, find the sofa, walk to sofa, sit in sofa, watch the television", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'crayon', 'chair', 'walllamp', 'knifeblock', 'bills', 'wallshelf', 'bathroom_counter', 'phone', 'ceiling', 'bedroom', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'dining_room', 'nightstand', 'oven', 'television', 'tray', 'curtain', 'towel_rack', 'door', 'bathtub', 'napkin', 'bag']\nTask:Read book\nresponse:\n{\n    \"thoughts\": walk to living room ,find a book ,open the book ,read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'character', 'stovefan', 'mat', 'tvstand', 'faucet', 'shower', 'clothes_pants', 'towel', 'wall', 'nightstand', 'powersocket', 'oven', 'light_bulb', 'tray', 'curtain', 'rag', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <towel>", "[GRAB] <towel>", "[CLOSE] <bathroom_cabinet>", "[FIND] <shower>", "[RINSE] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'wallshelf', 'cpuscreen', 'floor', 'shampoo', 'mop_bucket', 'toaster', 'mat', 'plate', 'shower', 'dish_soap', 'closetdrawer', 'dining_room', 'wall', 'photoframe', 'coffe_maker', 'colander', 'pencil', 'kitchen_counter', 'cupboard', 'clothes_dress', 'food_steak', 'home_office', 'dishwasher', 'soap', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'mat', 'ceilinglamp', 'tvstand', 'homework', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk to the office. I push the on button on the computer. I sit down in the chair. I work on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <button>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'chair', 'freezer', 'bathroom', 'trashcan', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'character', 'bedroom', 'computer', 'plate', 'stereo', 'window', 'food_kiwi', 'wall', 'nightstand', 'box', 'bed', 'clothes_gloves', 'coffe_maker', 'wooden_spoon', 'curtain', 'shoes', 'dishwasher', 'creditcard']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go into my bedroom and see my stereo. I decide to play music. I walk over to the stereo and turn it on. I sit down on the bed to listen to the music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[SWITCHON] <stereo>', '[FIND] <bed>', '[SIT] <bed>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "you go into your office, switch on the light, find chair at desk, sit in the chair, switch on the computer and work.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'orchid', 'hanger', 'ceiling', 'keyboard', 'piano_bench', 'tvstand', 'shower', 'electric_shaver', 'clothes_socks', 'couch', 'wall', 'powersocket', 'bench', 'cutting_board', 'toy', 'photoframe', 'bed', 'razor', 'wooden_spoon', 'after_shave', 'coffee_pot', 'dresser', 'clothes_dress', 'bathtub', 'bathroom_cabinet', 'light', 'clothes_shirt', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to kitchen. grab clothes_dress. look at washing machine. open washing machine. put clothes_dress on washing machine. close washing machine. switch on washing machine. switch off washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[TURNTO] <washing_machine>', '[LOOKAT] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>', '[SWITCHOFF] <washing_machine>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'measuring_cup', 'toilet', 'freezer', 'bathroom', 'cpuscreen', 'character', 'desk', 'bedroom', 'table_cloth', 'mat', 'shower', 'stereo', 'microphone', 'clothes_hat', 'pillow', 'nightstand', 'wall_clock', 'oven', 'keys', 'light_bulb', 'check', 'razor', 'tray', 'doorjamb', 'dresser', 'clothes_dress', 'bag', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I find a radio. I turn on the radio. I listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['cd_player', 'freezer', 'ground_coffee', 'bathroom', 'milk', 'orchid', 'cpuscreen', 'hanger', 'dvd_player', 'character', 'ceilinglamp', 'stereo', 'faucet', 'dustpan', 'clothes_hat', 'nightstand', 'wall', 'newspaper', 'check', 'mouthwash', 'video_game_controller', 'food_food', 'tray', 'doorjamb', 'pencil', 'cupboard', 'door', 'sink', 'home_office', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk in the living room, then I go the the stereo and the cd player and turn them on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['sponge', 'fork', 'pot', 'sheets', 'food_salt', 'detergent', 'food_butter', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'shampoo', 'desk', 'food_cake', 'mat', 'faucet', 'food_kiwi', 'wall', 'powersocket', 'toy', 'novel', 'food_food', 'curtain', 'dresser', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Going to the sink and scrubbing hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'bathroom', 'hanger', 'floor', 'phone', 'ceiling', 'character', 'bedroom', 'bookshelf', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'closetdrawer', 'couch', 'pillow', 'cup', 'coffe_maker', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'home_office', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'dough', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'bookmark', 'mat', 'faucet', 'shower', 'closetdrawer', 'couch', 'food_peanut_butter', 'microphone', 'pillow', 'towel', 'wall', 'food_food', 'television', 'kitchen_counter', 'cupboard', 'controller', 'door', 'bag', 'light', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "walk to kitchen and seprate the groceries and arrange in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'pencil', 'bathtub', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": For browsing internet walk to home office. Find Computer. Walk to Computer. Find Chair. Sit in a chair and switch on computer. Connect with internet and browse the net.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'microwave', 'chair', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'phone', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'stovefan', 'pillow', 'wall', 'oven', 'clothes_gloves', 'coffe_maker', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'instrument_guitar', 'dishwasher', 'soap']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walk into Kitchen, walk to sink, grab a plate,put soap on sponge,wash plate with sponge, rinse with water, put away plate in cupboard.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <plate>", "[FIND] <water>", "[RINSE] <water>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[PUTBACK] <plate> <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'mat', 'plate', 'faucet', 'ceilinglamp', 'dish_soap', 'mousepad', 'couch', 'dining_room', 'wall', 'oven', 'bench', 'basket_for_clothes', 'towel_rack', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'light']\nTask:Wash dirty dishes\nresponse:\n{\n    \"thoughts\": Wet the sponge, apply soap to sponge, scrub dishes, rinse dishes, place on dish rack.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <sponge> <sink>', '[GRAB] <sponge>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'band-aids', 'walllamp', 'nail_polish', 'drawing', 'bills', 'wallshelf', 'orchid', 'hanger', 'knife', 'floor', 'ceiling', 'desk', 'alcohol', 'mat', 'foundation', 'closetdrawer', 'couch', 'pillow', 'wall', 'bed', 'novel', 'check', 'oven_mitts', 'rag', 'kitchen_counter', 'dresser', 'bag', 'napkin']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I go to the book case, I select a book, I go back and sit down, then I read.", "actions": ["[WALK] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'toilet', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'ceilinglamp', 'tvstand', 'stereo', 'faucet', 'mousepad', 'dining_room', 'wall', 'powersocket', 'food_food', 'food_donut', 'towel_rack', 'mouse', 'shoes', 'dishrack', 'sink', 'spoon', 'light', 'coffee']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to bathroom, find toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'cat', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'cutting_board', 'coffee_table', 'novel', 'food_food', 'television', 'towel_rack', 'mouse', 'shoes', 'door', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I walk to the living room and pick up a book from the coffee table. I find my favorite chair.  I sit in the armchair and open the book.  I read.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[WALK] <coffee_table>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'chair', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'powersocket', 'television', 'food_rice', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'door', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": turn on the light for room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[SWITCHON] <light>', '[FIND] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'food_salt', 'chair', 'freezer', 'cat', 'orchid', 'knife', 'bathroom_counter', 'ceiling', 'folder', 'purse', 'bookmark', 'ceilinglamp', 'stereo', 'closetdrawer', 'clothes_pants', 'dirt', 'wall', 'food_food', 'wine', 'rag', 'console', 'shaving_cream', 'pencil', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'home_office', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk into laundry room. Open washing machine door. Put dirty clothes into washing machine. Measure laundry detergent and place into washing machine. Turn dial to appropriate wash setting. Power on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[CLOSE] <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[OPEN] <washing_machine>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'pot', 'freezer', 'milk', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'alcohol', 'computer', 'toaster', 'shoe-shine_kit', 'faucet', 'pillow', 'dining_room', 'wall', 'cup', 'bed', 'check', 'toilet_paper', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'clothes_dress', 'food_steak', 'sink', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk into the kitchen. I reach into the cabinet and pick up a water glass. Turn on the faucet and place the glass under the water stream. Bring the glass of water to lips and drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'tea', 'microwave', 'freezer', 'ground_coffee', 'hanger', 'floor', 'ceiling', 'dvd_player', 'bedroom', 'computer', 'toaster', 'shoe-shine_kit', 'shower', 'picture', 'wall', 'keys', 'coffe_maker', 'check', 'video_game_controller', 'coffee_filter', 'oven_mitts', 'water', 'television', 'after_shave', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'window', 'toaster', 'clothes_pants', 'broom', 'powersocket', 'bench', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'dresser', 'hair', 'door', 'sink', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Get some clothes together that make an appropriate outfit.  Put them on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'cat', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'ceilinglamp', 'brush', 'faucet', 'stereo', 'tvstand', 'towel', 'pillow', 'wall', 'pasta', 'powersocket', 'coffe_maker', 'oven_mitts', 'tray', 'curtain', 'dresser', 'cupboard', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[PUTOBJBACK] <soap>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'beer', 'trashcan', 'dough', 'cpuscreen', 'desk', 'keyboard', 'toaster', 'stovefan', 'tvstand', 'shower', 'faucet', 'window', 'oil', 'couch', 'dining_room', 'cutting_board', 'bed', 'food_food', 'television', 'ice', 'doorjamb', 'towel_rack', 'cd', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I hear the phone ringing in the living room. I walk to the living room and go to the phone. I pick it up and answer the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>', '[SWITCHON] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'novel', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'comb', 'door', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Find book, grab book, fin bookmark, sit down somewhere comfortable, open book, read book, close book after reading.", "actions": ["[WALK] <desk>", "[FIND] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[FIND] <couch>", "[SIT] <couch>", "[GRAB] <novel>", "[READ] <novel>", "[PUTBACK] <bookmark> <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'bookmark', 'stovefan', 'ceilinglamp', 'mat', 'mousepad', 'oil', 'food_peanut_butter', 'broom', 'wall', 'novel', 'diary', 'mouse', 'cupboard', 'clothes_dress', 'door', 'table', 'light', 'dirt']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Go to office. Turn on computer. Pull out chair. Sit in chair. Use keyboard and mouse.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[GRAB] <mouse>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'cards', 'dresser', 'door', 'colander']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'sheets', 'filing_cabinet', 'food_bread', 'food_salt', 'microwave', 'chair', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'mail', 'stovefan', 'ceilinglamp', 'mat', 'stereo', 'couch', 'wall', 'powersocket', 'cup', 'check', 'mouthwash', 'food_donut', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'table', 'teeth']\nTask:Play games\nresponse:\n{\n    \"thoughts\": I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <board_game>', '[FIND] <board_game>', '[GRAB] <board_game>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <board_game> <table>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mop_bucket', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "its time for my favorite show and im going change with the help of remote control and watch", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <chair>", "[SIT] <chair>", "[TOUCH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'detergent', 'ceilingfan', 'drawing', 'needle', 'knifeblock', 'floor', 'ceiling', 'phone', 'bedroom', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'bench', 'check', 'television', 'oven_mitts', 'curtain', 'conditioner', 'food_steak', 'home_office', 'table', 'vase', 'clothes_shirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the sofa to watch tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'measuring_cup', 'toilet', 'drawing', 'needle', 'hanger', 'floor', 'computer', 'mat', 'mousepad', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'bed', 'novel', 'food_food', 'video_game_controller', 'coffe_maker', 'tray', 'food_onion', 'doorjamb', 'lighter', 'cd', 'clothes_dress', 'bathroom_cabinet', 'light', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I would go lie down in my bed and open the book and start reading.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'food_butter', 'floor', 'ceiling', 'stovefan', 'faucet', 'dustpan', 'food_jam', 'wall', 'oven', 'cutting_board', 'newspaper', 'coffe_maker', 'water', 'drinking_glass', 'tray', 'doorjamb', 'kitchen_counter', 'vacuum_cleaner', 'pajamas', 'home_office', 'bathroom_cabinet', 'clothes_skirt', 'vase', 'scrabble']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to the kitchen,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <water>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'tea_bag', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'hair', 'door']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I walk to the Home Office then find the computer and sit in the chair. I switch on the computer and start typing the email. I switch the computer off.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'toaster', 'mat', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'bench', 'photoframe', 'coffe_maker', 'food_rice', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'dishwasher', 'home_office']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'blender', 'toilet', 'jelly', 'chair', 'drawing', 'bathroom_counter', 'floor', 'character', 'food_bacon', 'mat', 'shoe-shine_kit', 'shower', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'pasta', 'oven', 'newspaper', 'check', 'television', 'ice', 'tray', 'doorjamb', 'pencil', 'comb', 'dresser', 'bathroom_cabinet']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom. Pull down pants. Use the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[PULL] <clothes_pants>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'toy', 'photoframe', 'bed', 'television', 'doorjamb', 'towel_rack', 'dresser', 'dishwasher', 'shoe_rack']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Sit down on couch in front of TV.  Use remote to turn on the TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'juice', 'toilet', 'freezer', 'bathroom', 'bills', 'orchid', 'shampoo', 'ceiling', 'stamp', 'desk', 'keyboard', 'tvstand', 'couch', 'dustpan', 'dining_room', 'powersocket', 'newspaper', 'check', 'video_game_controller', 'television', 'coffe_maker', 'ice', 'tray', 'doorjamb', 'shoes', 'controller', 'door', 'napkin']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find radio, turn it on", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'chair', 'drawing', 'needle', 'hanger', 'ceiling', 'phone', 'toaster', 'ceilinglamp', 'laptop', 'iron', 'mat', 'deck_of_cards', 'mousepad', 'oil', 'wall', 'basket_for_clothes', 'food_donut', 'food_rice', 'tray', 'curtain', 'doorjamb', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'light']\nTask:Put on coat\nresponse:\n{\n    \"thoughts\": I remove my scarf from a coat peg in the entrance hall.  I wrap the scarf around my neck,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <hanger>', '[TURNTO] <hanger>', '[LOOKAT] <hanger>', '[FIND] <clothes_scarf>', '[GRAB] <clothes_scarf>', '[PUTON] <clothes_scarf>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'tablelamp', 'toilet', 'freezer', 'detergent', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'stovefan', 'mat', 'microphone', 'pillow', 'nightstand', 'wall', 'powersocket', 'bed', 'novel', 'television', 'doorjamb', 'headset', 'shoes', 'controller', 'door', 'light', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the bedroom. Grab the book out of the nightstand. Lay down in bed. Open the book. Read the book.", "actions": ["[WALK] <bedroom>", "[WALK] <nightstand>", "[FIND] <nightstand>", "[OPEN] <nightstand>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[TURNTO] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'freezer', 'bathroom', 'food_carrot', 'knifeblock', 'food_egg', 'hanger', 'floor', 'phone', 'stovefan', 'mat', 'plate', 'stereo', 'dish_soap', 'food_noodles', 'closetdrawer', 'mousepad', 'dustpan', 'cup', 'food_food', 'coffe_maker', 'check', 'oven_mitts', 'water', 'food_rice', 'curtain', 'mouse', 'hairbrush', 'sink', 'home_office']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": walk to kitchen, fins soap, arrange dishes, wash dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[FIND] <dish_soap>', '[FIND] <water>', '[GRAB] <plate>', '[GRAB] <dish_soap>', '[WASH] <plate>', '[RINSE] <plate>', '[DROP] <plate>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'chair', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'powersocket', 'television', 'food_rice', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "turn on the light for room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[SWITCHON] <light>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'phone', 'bookshelf', 'table_cloth', 'bookmark', 'toaster', 'mat', 'faucet', 'oven', 'cup', 'photoframe', 'check', 'television', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'door', 'sink', 'light', 'tooth_paste']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'bed', 'check', 'mouthwash', 'television', 'doorjamb', 'dresser', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room. look at remote control. grab remote control. plug in television. switch on television. lie in sofa. point at remote control. watch television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[LIE] <couch>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'fork', 'filing_cabinet', 'tablelamp', 'chair', 'band-aids', 'floor', 'ceiling', 'stamp', 'alcohol', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'towel', 'wall', 'toy', 'bed', 'keys', 'photoframe', 'coffe_maker', 'television', 'coin', 'kitchen_counter', 'dresser', 'cupboard', 'food_steak', 'sink', 'table']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Run warm water in the sink. Pour dish washing liquid on the dishes. Pick up dishes and wash them individually with a sponge. Rinse suds off under running water. Dry dishes with towel.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[WASH] <plate>', '[RINSE] <plate>', '[RINSE] <sponge>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <sponge>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <plate>', '[PUTOBJBACK] <towel>', '[WALK] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'microwave', 'chair', 'freezer', 'food_carrot', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'clothes_jacket', 'table_cloth', 'mat', 'tvstand', 'shower', 'stereo', 'couch', 'pillow', 'wall', 'oven', 'cup', 'food_food', 'television', 'coffee_filter', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'food_steak', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk Into Kitchen. Open Refrigerator. Grab Food. Walk to Microwave. Insert food into microwave. Cook until warm.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[CLOSE] <freezer>", "[WALK] <microwave>", "[SWITCHON] <microwave>", "[SWITCHOFF] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'coin', 'curtain', 'doorjamb', 'coffee_pot', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'knife', 'wallshelf', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'box', 'photoframe', 'television', 'coffee_filter', 'after_shave', 'doorjamb', 'diary', 'kitchen_counter', 'dresser']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to livingroom, grab remote and change the channel on television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'freezer', 'walllamp', 'knifeblock', 'wallshelf', 'clothes_underwear', 'ceiling', 'bookshelf', 'keyboard', 'ceilinglamp', 'plate', 'stereo', 'clothes_socks', 'clothes_pants', 'pillow', 'bowl', 'powersocket', 'cutting_board', 'basket_for_clothes', 'razor', 'coin', 'wine', 'curtain', 'pencil', 'kitchen_counter', 'cd', 'dresser', 'laundry_detergent', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'oil', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'creditcard']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to the Home office. Sit on the chair. Switch on the computer. Type with keyboard for internet.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'remote_control', 'computer', 'button', 'toaster', 'mat', 'shower', 'food_peanut_butter', 'food_kiwi', 'nightstand', 'oven', 'photoframe', 'television', 'oven_mitts', 'tray', 'kettle', 'cards', 'pencil', 'kitchen_counter', 'towel_rack', 'bathtub', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk into living room. Find television remote control. Pick up remote control. Press button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'soap']\nTask:Work\nresponse:\n", "output": {"thoughts": "I go to my home office, sit on the chair and turn on the computer. I open my email account to look for all the information I need.I pull up my online task on the computer to look at what the assigned task is.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'beer', 'microwave', 'freezer', 'walllamp', 'dough', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'oven', 'cutting_board', 'shelf', 'coffe_maker', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'cards', 'kitchen_counter', 'maindoor', 'bathtub', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Get a can of beer, open it, sit on couch, relax,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <beer>', '[GRAB] <beer>', '[CLOSE] <freezer>', '[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[DRINK] <beer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'detergent', 'drawing', 'knife', 'wallshelf', 'hanger', 'floor', 'ceiling', 'computer', 'mop_bucket', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'couch', 'clothes_hat', 'pillow', 'wall', 'box', 'photoframe', 'bed', 'coffe_maker', 'television', 'coin', 'doorjamb', 'door', 'sink']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room ,find sofa ,sit on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'foundation', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'napkin', 'light', 'table', 'dirt']\nTask:Work\nresponse:\n{\n    \"thoughts\": go to home office. turn the computer on. begin working,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'tablelamp', 'chair', 'cpuscreen', 'floor', 'ceiling', 'character', 'keyboard', 'computer', 'toaster', 'mat', 'stovefan', 'faucet', 'shower', 'mousepad', 'broom', 'wall', 'powersocket', 'cup', 'novel', 'coffe_maker', 'mouthwash', 'tray', 'kitchen_counter', 'cupboard', 'conditioner', 'sink', 'napkin', 'soap']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drink water from a cup", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[FIND] <faucet>", "[WALK] <sink>", "[PUTBACK] <cup> <sink>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[GRAB] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'shampoo', 'stovefan', 'mat', 'shower', 'window', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'newspaper', 'food_food', 'drinking_glass', 'water', 'curtain', 'doorjamb', 'dresser', 'food_steak', 'door', 'home_office', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'homework', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'keys', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "walk to Office,Walk to desk ,sit down in chair, turn on computer, start homework.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <check>", "[FIND] <homework>", "[TURNTO] <homework>", "[LOOKAT] <homework>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'trashcan', 'bills', 'knife', 'floor', 'desk', 'mail', 'button', 'toaster', 'plate', 'faucet', 'shower', 'foundation', 'couch', 'bowl', 'wall', 'nightstand', 'oven', 'bed', 'clothes_gloves', 'novel', 'coffe_maker', 'check', 'tray', 'sink', 'table', 'dishwasher', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['cd_player', 'freezer', 'ground_coffee', 'bathroom', 'milk', 'orchid', 'cpuscreen', 'hanger', 'dvd_player', 'character', 'ceilinglamp', 'stereo', 'faucet', 'dustpan', 'clothes_hat', 'nightstand', 'wall', 'newspaper', 'check', 'mouthwash', 'video_game_controller', 'food_food', 'tray', 'doorjamb', 'pencil', 'cupboard', 'door', 'sink', 'home_office', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk in the living room, then I go the the stereo and the cd player and turn them on.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'powersocket', 'oven', 'coffe_maker', 'television', 'coffee_filter', 'food_rice', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'dirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I like to set on my sofa.Turn on the tv. And watch Perry Mason. Monday's through Friday's.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'dog', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'mousepad', 'wall', 'nightstand', 'oven', 'photoframe', 'novel', 'television', 'lighter', 'towel_rack', 'mouse', 'dresser', 'door']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Reading a book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'drawing', 'orchid', 'wallshelf', 'floor', 'table_cloth', 'toaster', 'ceilinglamp', 'plate', 'shower', 'faucet', 'dish_soap', 'closetdrawer', 'pillow', 'wall', 'novel', 'coffe_maker', 'food_rice', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'clothes_scarf', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find dishwasher. Open dishwasher. Find dishrack. Find plate 1. Grab plate 1. Put plate 1 in dishrack. Repeat for dish 2. Find dish soap. Grab dish soap. Put dish soap to dishwasher. Close dishwasher. Switch on dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'button', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'iron', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'newspaper', 'food_food', 'television', 'doorjamb', 'dresser', 'conditioner']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk into the Living room.  Find the television remote.  Use it to turn on the television.  Point the remote at the television. Press the channel button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'walllamp', 'milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'candle', 'pillow', 'wall', 'cup', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'bathroom_cabinet', 'teeth']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'water_glass', 'food_salt', 'freezer', 'walllamp', 'blow_dryer', 'detergent', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'stovefan', 'mat', 'iron', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'newspaper', 'keys', 'food_food', 'television', 'kitchen_counter', 'cupboard', 'door', 'bathroom_cabinet', 'tooth_paste']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen ,find fridge ,open fridge ,find water  glass , drink from water glass", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'television', 'tray', 'curtain', 'dresser', 'bathtub', 'sink', 'bathroom_cabinet', 'coffee_cup', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'mouse', 'dresser', 'conditioner', 'door', 'table', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "For relaxing walk to living room. Find the sofa to relax. Arrange the sofa to relax. Sit on the sofa to relaxing.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <conditioner>", "[FIND] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'coffe_maker', 'television', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'home_office', 'table', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the living room and find the book you want to read. Take it and then go to the couch. Sit on it, open the book and start reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'food_butter', 'drawing', 'knife', 'bathroom_counter', 'ceiling', 'phone', 'desk', 'bedroom', 'clothes_jacket', 'faucet', 'shower', 'mousepad', 'couch', 'clothes_pants', 'wall', 'powersocket', 'food_donut', 'tray', 'curtain', 'after_shave', 'pencil', 'vacuum_cleaner', 'clothes_dress', 'sink', 'napkin', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open cabinet and collect clothes and put them into washing machine and wash clothes", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['trashcan', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'folder', 'keyboard', 'remote_control', 'alcohol', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'couch', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'curtain', 'food_steak', 'napkin', 'table', 'clothes_shirt', 'creditcard']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Take TV remote to change TV Channel. Press channel/program +/- button on TV remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[TOUCH] <button>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'tablelamp', 'microwave', 'hanger', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'shower', 'mousepad', 'couch', 'candle', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'drinking_glass', 'wooden_spoon', 'curtain', 'tray', 'doorjamb', 'kitchen_counter', 'paper_towel', 'cupboard', 'sink', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "wipe loose dirt, place in the dishwasher, turn on the dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WIPE] <drinking_glass>", "[RINSE] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'blender', 'microwave', 'toilet', 'bathroom', 'food_carrot', 'milk', 'wallshelf', 'cpuscreen', 'toaster', 'window', 'clothes_socks', 'pillow', 'wall', 'standingmirror', 'food_food', 'mouthwash', 'razor', 'ice', 'rag', 'lighter', 'cards', 'mouse', 'cd', 'hairbrush', 'pajamas', 'dresser', 'laundry_detergent', 'sink', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": walk to living room ,find a book ,open the book ,read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'milk', 'toothbrush', 'drawing', 'orchid', 'cpuscreen', 'character', 'bedroom', 'stovefan', 'toaster', 'mousepad', 'clothes_pants', 'food_kiwi', 'wall_clock', 'bed', 'television', 'toothbrush_holder', 'water', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'shoes', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'dirt']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'microwave', 'chair', 'jelly', 'bathroom', 'detergent', 'knife', 'floor', 'toaster', 'mat', 'plate', 'shower', 'faucet', 'dish_soap', 'bowl', 'wall', 'cup', 'coffe_maker', 'food_food', 'after_shave', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'spoon', 'home_office', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the dishwasher. I open it. I find the sink and grab the dirty plates, cups, bowls, and utensils. I rinse off the dirty dishes under the faucet. I then place the dirty dishes into the dishwasher. I add soap to the dishwasher and close the lid. I turn it on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <sink>', '[FIND] <bowl>', '[GRAB] <bowl>', '[RINSE] <bowl>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[RINSE] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[RINSE] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[RINSE] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[RINSE] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[TURNTO] <dishwasher>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'crayon', 'microwave', 'chair', 'orchid', 'drawing', 'wallshelf', 'floor', 'character', 'table_cloth', 'ceilinglamp', 'plate', 'stereo', 'mat', 'tvstand', 'wall', 'oven', 'cup', 'bed', 'food_food', 'razor', 'oven_mitts', 'tray', 'kitchen_counter', 'spoon', 'table', 'dishwasher', 'teeth', 'creditcard']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Place your plate on the placemat. The forks will be placed on the left side of the plate and the knife and spoon on the right and place the spoon to the right of the knife. If you want to set out food, place them above the plate.", "actions": ["[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[DROP] <plate>", "[FIND] <fork>", "[GRAB] <fork>", "[DROP] <fork>", "[FIND] <spoon>", "[GRAB] <spoon>", "[DROP] <spoon>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'laptop', 'plate', 'stovefan', 'shower', 'dish_soap', 'wall', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'kettle', 'pencil', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher', 'coffee']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put the dishes on dinner table in the dishwasher and wash them,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[WALK] <table>', '[FIND] <pot>', '[GRAB] <pot>', '[WALK] <dishwasher>', '[PUTBACK] <pot> <dishwasher>', '[CLOSE] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[OPEN] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'box', 'novel', 'toilet_paper', 'television', 'mouse', 'dresser', 'door', 'food_steak', 'clothes_skirt', 'tooth_paste']\nTask:Read book\nresponse:\n", "output": {"thoughts": "\"walk to entrance\" \"walk to living room\" \"look for book\" \"find book\" \"pick up book\" \"look for chair\" \"find chair\" \"sit in chair\" \"read book\"", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[FIND] <novel>", "[FIND] <chair>", "[TURNTO] <chair>", "[LOOKAT] <chair>", "[SIT] <chair>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'mop', 'walllamp', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'remote_control', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'powersocket', 'box', 'love_seat', 'photoframe', 'television', 'curtain', 'doorjamb', 'dresser', 'clothes_dress', 'bathtub', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I'll watch TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'microwave', 'chair', 'walllamp', 'cat', 'wallshelf', 'floor', 'character', 'desk', 'tea_bag', 'toaster', 'plate', 'couch', 'dining_room', 'box', 'newspaper', 'bed', 'toilet_paper', 'curtain', 'coffee_pot', 'mouse', 'kettle', 'cupboard', 'controller', 'clothes_dress', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'soap']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find Cat in living room.  Walk to cat.  Pet cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'ground_coffee', 'drawing', 'orchid', 'needle', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'wall', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'hair', 'clothes_dress', 'door', 'table', 'clothes_skirt', 'scrabble']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'mop', 'dog', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'toaster', 'mat', 'closetdrawer', 'couch', 'food_jam', 'food_orange', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'food_cheese', 'keys', 'food_food', 'video_game_controller', 'television', 'ice', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Open fridge. Put groceries in fridge.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'chair', 'bathroom', 'food_carrot', 'knifeblock', 'floor', 'phone', 'bedroom', 'envelope', 'mail', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'tvstand', 'shower', 'window', 'candle', 'microphone', 'pillow', 'basket_for_clothes', 'toilet_paper', 'check', 'rag', 'chessboard', 'laundry_detergent', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Pick up my dirty clothes, load the washer, add detergent, turn on washer,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[RUN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'blender', 'toilet', 'bathroom', 'detergent', 'food_butter', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'mousepad', 'couch', 'food_kiwi', 'wall', 'oven', 'toy', 'coffe_maker', 'television', 'oven_mitts', 'coffee_filter', 'ice', 'coffee_pot', 'shaving_cream', 'kitchen_counter', 'cd', 'cupboard', 'sink', 'home_office', 'dishwasher', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'chair', 'bathroom', 'drawing', 'food_egg', 'floor', 'desk', 'bookshelf', 'computer', 'mat', 'stereo', 'shower', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'cup', 'doorjamb', 'mouse', 'kitchen_counter', 'pajamas', 'controller', 'door', 'clothes_scarf', 'light', 'dishwasher', 'board_game']\nTask:Pull up carpet\nresponse:\n{\n    \"thoughts\": Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mat>', '[FIND] <mat>', '[TOUCH] <mat>', '[PULL] <mat>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'chair', 'freezer', 'walllamp', 'detergent', 'orchid', 'phone', 'bookshelf', 'computer', 'toaster', 'tvstand', 'stereo', 'closetdrawer', 'food_peanut_butter', 'clothes_pants', 'form', 'wall', 'oven', 'cup', 'bed', 'mouthwash', 'rag', 'towel_rack', 'kitchen_counter', 'pajamas', 'food_steak', 'door', 'napkin', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the playback device you want to listen to. turn on. press play. enjoyu music.", "actions": ["[WALK] <bedroom>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_salt', 'microwave', 'freezer', 'orchid', 'cpuscreen', 'ceiling', 'character', 'bedroom', 'desk', 'food_bacon', 'bookmark', 'toaster', 'stovefan', 'faucet', 'shower', 'mousepad', 'wall', 'wall_clock', 'oven', 'bench', 'coffe_maker', 'television', 'water', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen, take a glass from the cupboard, fill with water and drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>', '[RINSE] <water_glass>', '[PUTBACK] <water_glass> <cupboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'toilet_paper', 'television', 'doorjamb', 'dresser', 'cupboard', 'light', 'clothes_shirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Go to the living room, Switch on the television. Find the remote and press the change channel button, change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[TOUCH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'cpuscreen', 'clothes_underwear', 'hanger', 'shampoo', 'bookshelf', 'alcohol', 'table_cloth', 'faucet', 'shower', 'clothes_socks', 'clothes_pants', 'wall', 'wall_clock', 'oven', 'nightstand', 'basket_for_clothes', 'newspaper', 'clothes_gloves', 'tray', 'doorjamb', 'lighter', 'cards', 'clothes_dress', 'laundry_detergent', 'napkin', 'clothes_skirt', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bathroom, take the dirt clothes and put into the washing machine, then put the soap, set the cleaning program and switch on the machine,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_skirt>', '[GRAB] <clothes_skirt>', '[PUTBACK] <clothes_skirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[CLOSE] <basket_for_clothes>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'food_salt', 'mop', 'toilet', 'freezer', 'bathroom', 'dough', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'mat', 'ceilinglamp', 'couch', 'food_orange', 'pillow', 'wall', 'oven', 'tape', 'food_cheese', 'food_food', 'coffe_maker', 'television', 'tray', 'kitchen_counter', 'dishrack', 'cupboard', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Put groceries into the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'filing_cabinet', 'juice', 'milk', 'hanger', 'kitchen_cabinet', 'bathroom_counter', 'button', 'food_bacon', 'stovefan', 'mat', 'shoe-shine_kit', 'tvstand', 'electric_shaver', 'food_peanut_butter', 'form', 'dining_room', 'cutting_board', 'cup', 'photoframe', 'coffe_maker', 'television', 'water', 'wooden_spoon', 'kettle', 'kitchen_counter', 'chessboard', 'dresser', 'sink', 'spoon']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Go to kitchen. Open the the cabinet. Locate the water glass. Pour water into glass. Drag the bottle close to mouth. Drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[CLOSE] <kitchen_cabinet>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'wall', 'pasta', 'oven', 'powersocket', 'box', 'water', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'sauce_pan', 'cupboard', 'door', 'light', 'dishwasher', 'tooth_paste', 'teeth']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Cook some pasta on the stove", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pasta>", "[GRAB] <pasta>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <sauce_pan>", "[POUR] <water> <sauce_pan>", "[FIND] <oven>", "[SWITCHON] <oven>", "[PUTBACK] <pasta> <sauce_pan>", "[FIND] <sauce_pan>", "[GRAB] <sauce_pan>", "[PUTBACK] <sauce_pan> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'juice', 'microwave', 'freezer', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'tea_bag', 'keyboard', 'computer', 'button', 'stovefan', 'mat', 'ceilinglamp', 'homework', 'closetdrawer', 'couch', 'wall', 'oven', 'cutting_board', 'toilet_paper', 'television', 'curtain', 'diary', 'pencil', 'shoes', 'dresser', 'table']\nTask:Put out flowers\nresponse:\n{\n    \"thoughts\": Place a flower centrepiece on the entrance hall table.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <centerpiece>', '[FIND] <centerpiece>', '[GRAB] <centerpiece>', '[WALK] <table>', '[PUTBACK] <centerpiece> <table>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'towel', 'dining_room', 'wall', 'nightstand', 'cup', 'photoframe', 'television', 'food_rice', 'doorjamb', 'mouse', 'dresser', 'clothes_scarf', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the living room. Sit on the sofa. Grab the remote and turn on the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'keyboard', 'remote_control', 'bookshelf', 'computer', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'broom', 'nightstand', 'wall', 'television', 'curtain', 'doorjamb', 'pencil', 'dresser', 'pajamas', 'laundry_detergent', 'clothes_dress', 'bag', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk into the living room walk to the remote pick it up and turn in the tv and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'controller', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I need to finish some work on my computer in my home office to meet a deadline for tomorrow.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <desk>", "[OPEN] <desk>", "[PUTBACK] <keyboard> <desk>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <mousepad>", "[PUTBACK] <mouse> <mousepad>", "[PLUGIN] <keyboard>", "[PLUGIN] <mouse>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[TYPE] <keyboard>", "[GRAB] <mouse>", "[TOUCH] <mouse>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'trashcan', 'drawing', 'knife', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'check', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'controller', 'bathtub', 'home_office', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find cat. Walk to cat. Pull cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[PULL] <cat>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'ground_coffee', 'cat', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'food_bacon', 'mat', 'iron', 'couch', 'food_peanut_butter', 'food_orange', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'clothes_gloves', 'food_food', 'television', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'hair', 'clothes_dress', 'light', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I put my groceries into the fridge.", "actions": ["[WALK] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'freezer', 'orchid', 'floor', 'bedroom', 'keyboard', 'stovefan', 'stereo', 'shower', 'closetdrawer', 'couch', 'dining_room', 'tape', 'cup', 'bed', 'novel', 'television', 'toothbrush_holder', 'doorjamb', 'mouse', 'kitchen_counter', 'chessboard', 'cupboard', 'door', 'spoon', 'instrument_guitar', 'table', 'clothes_shirt', 'shoe_rack']\nTask:Look out window\nresponse:\n{\n    \"thoughts\": I head into my child's room and look out the window to view the neighborhood from a different area in the house.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <window>', '[TURNTO] <window>', '[LOOKAT] <window>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'television', 'doorjamb', 'mouse', 'pencil', 'dresser', 'door', 'dishwasher']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to home office and read a book", "actions": ["[RUN] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[FIND] <floor>", "[PUTBACK] <bookmark> <floor>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'fork', 'filing_cabinet', 'tablelamp', 'chair', 'band-aids', 'floor', 'ceiling', 'stamp', 'alcohol', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'towel', 'wall', 'toy', 'bed', 'keys', 'photoframe', 'coffe_maker', 'television', 'coin', 'kitchen_counter', 'dresser', 'cupboard', 'food_steak', 'sink', 'table']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Run warm water in the sink. Pour dish washing liquid on the dishes. Pick up dishes and wash them individually with a sponge. Rinse suds off under running water. Dry dishes with towel.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[WASH] <plate>', '[RINSE] <plate>', '[RINSE] <sponge>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <sponge>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <plate>', '[PUTOBJBACK] <towel>', '[WALK] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'filing_cabinet', 'microwave', 'walllamp', 'drawing', 'food_egg', 'wallshelf', 'floor', 'stamp', 'computer', 'alcohol', 'toaster', 'plate', 'shower', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'water', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'spoon', 'instrument_guitar', 'light', 'dishwasher', 'soap', 'scrabble']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "first i will take all the dishes to the dish washer and with the help of dishwasher and water i will clean all the dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water>", "[TURNTO] <water>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_pizza', 'chair', 'drawing', 'food_egg', 'hanger', 'floor', 'phone', 'ceiling', 'mail', 'table_cloth', 'stovefan', 'ceilinglamp', 'stereo', 'homework', 'couch', 'microphone', 'form', 'food_orange', 'bowl', 'oven', 'box', 'newspaper', 'tray', 'kitchen_counter', 'cupboard', 'controller', 'bathroom_cabinet', 'table', 'shoe_rack']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to table, grab water glass and drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>', '[PUTBACK] <water_glass> <table>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'freezer', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'toaster', 'mat', 'faucet', 'stereo', 'food_jam', 'food_orange', 'pillow', 'towel', 'wall', 'food_cheese', 'coffe_maker', 'check', 'coffee_filter', 'water', 'drinking_glass', 'curtain', 'kitchen_counter', 'shoes', 'cupboard', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Walk to faucet. Switch on faucet. Pour water into drinking glass. Switch off faucet. Drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <sink>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[SWITCHOFF] <faucet>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'pot', 'tablelamp', 'measuring_cup', 'freezer', 'detergent', 'cat', 'bathroom_counter', 'keyboard', 'purse', 'bookshelf', 'toaster', 'brush', 'faucet', 'stereo', 'homework', 'toy', 'box', 'basket_for_clothes', 'coin', 'curtain', 'rag', 'centerpiece', 'dishrack', 'clothes_dress', 'bag', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the washing machine. Open the washing machine. Find the basket for clothes. Find the clothes_dress. Grab the clothes_dress. Put the clothes_dress into the washing machine. Find the clothes_shirt. Grab the clothes_shirt. Put the clothes_shirt into the washing machine. Close the washing machine. Switch on the washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <basket_for_clothes>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'water_glass', 'food_ice_cream', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'toaster', 'stovefan', 'mat', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'wall', 'oven', 'powersocket', 'cutting_board', 'coffe_maker', 'television', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'hair', 'door', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I enter the kitchen and grab a waterglass from the cupboard. I open the fridge and grap some water. I pour the water into the cup and drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[OPEN] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'tea_bag', 'computer', 'deck_of_cards', 'mousepad', 'couch', 'pillow', 'wall', 'child', 'bed', 'coin', 'coffee_filter', 'doorjamb', 'centerpiece', 'door', 'table', 'light', 'dishwasher']\nTask:Let baby learn how to walk\nresponse:\n{\n    \"thoughts\": The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[GREET] <child>', '[FIND] <couch>', '[TURNTO] <couch>', '[POINTAT] <couch>', '[TURNTO] <child>', '[LOOKAT] <child>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'blow_dryer', 'nail_polish', 'ground_coffee', 'milk', 'knife', 'wallshelf', 'hanger', 'character', 'desk', 'keyboard', 'stovefan', 'mat', 'plate', 'food_jam', 'nightstand', 'wall', 'toy', 'bed', 'coffe_maker', 'wine', 'rag', 'centerpiece', 'pencil', 'dresser', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up the book, open to current page, read the words.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'sheets', 'food_ice_cream', 'food_salt', 'measuring_cup', 'microwave', 'toilet', 'detergent', 'ceilingfan', 'drawing', 'hanger', 'floor', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'plate', 'window', 'floor_lamp', 'mousepad', 'couch', 'pillow', 'wall', 'clothes_gloves', 'novel', 'check', 'television', 'food_rice', 'doorjamb', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'tea', 'ground_coffee', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'button', 'mat', 'deck_of_cards', 'plate', 'tvstand', 'closetdrawer', 'couch', 'food_jam', 'dining_room', 'wall', 'powersocket', 'basket_for_clothes', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'chessboard', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk into living room, walk to light, turn on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'detergent', 'food_butter', 'toothbrush', 'drawing', 'floor', 'ceiling', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'dish_soap', 'wall', 'nightstand', 'cup', 'coffe_maker', 'coffee_filter', 'drinking_glass', 'tray', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'sink', 'coffee_cup', 'dishwasher', 'shoe_rack']\nTask:Clean dishes\nresponse:\n{\n    \"thoughts\": Wash dishes in the sink then place them in the dishwasher. Put detergent in the dishwasher and turn it on when full.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <drinking_glass>', '[WASH] <drinking_glass>', '[FIND] <coffee_cup>', '[WASH] <coffee_cup>', '[FIND] <plate>', '[WASH] <plate>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <dishwasher>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <detergent>', '[GRAB] <detergent>', '[PUTBACK] <detergent> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'nail_polish', 'cat', 'drawing', 'knife', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'photoframe', 'toilet_paper', 'check', 'television', 'doorjamb', 'towel_rack', 'mouse', 'vacuum_cleaner', 'dresser', 'cupboard', 'home_office']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'mop', 'microwave', 'walllamp', 'detergent', 'food_carrot', 'knifeblock', 'wallshelf', 'hanger', 'floor', 'ceiling', 'keyboard', 'stovefan', 'toaster', 'stereo', 'window', 'oil', 'clothes_socks', 'wall', 'oven', 'food_vegetable', 'bed', 'photoframe', 'television', 'food_donut', 'water', 'tray', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": walk to kicthen , find stove , find cooking pot , find vegetables , find water  ,put vegetables in cooking pot, pour water to cooking pot,put cooking pot on stove, switch on stove,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <oven>', '[FIND] <pot>', '[FIND] <food_vegetable>', '[FIND] <water>', '[GRAB] <water>', '[WASH] <food_vegetable>', '[POUR] <water> <pot>', '[FIND] <pot>', '[GRAB] <pot>', '[PUTBACK] <pot> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'jelly', 'walllamp', 'drawing', 'floor', 'ceiling', 'computer', 'stovefan', 'mat', 'shower', 'microphone', 'wall', 'oven', 'cup', 'photoframe', 'coffe_maker', 'television', 'tray', 'curtain', 'wooden_spoon', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'table', 'dishwasher', 'board_game', 'dirt']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to the bathroom and wash with the shower.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[WASH] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'sheets', 'pot', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mail', 'tvstand', 'stereo', 'window', 'mousepad', 'clothes_socks', 'wall', 'nightstand', 'powersocket', 'pasta', 'oven', 'light_bulb', 'coffe_maker', 'food_rice', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'door', 'light']\nTask:Turn light off\nresponse:\n{\n    \"thoughts\": Put the light switch into the down, or off, position.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Find the computer.  Turn on the computer.  Open the e-mail program.  Create a new message.  Write the message.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'crayon', 'microwave', 'toilet', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'shampoo', 'computer', 'mat', 'tvstand', 'picture', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'bathroom_cabinet', 'light', 'instrument_violin']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk over to the phone. .I reach my arm out. I open my hand. I place my hand on the phone. I close my fingers around the phone. I pick up the phone.,\n    \"actions\": ['[WALK] <phone>', '[TOUCH] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'mop', 'chair', 'detergent', 'cat', 'phone', 'stamp', 'keyboard', 'stovefan', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'deck_of_cards', 'toaster', 'mousepad', 'closetdrawer', 'dining_room', 'food_kiwi', 'wall', 'check', 'coffe_maker', 'curtain', 'rag', 'mouse', 'comb', 'bathroom_cabinet', 'dishwasher', 'soap']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go looking for the cat in the living room where he usually sleeps. Once I find the cat I just pet him for a few minutes.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'food_salt', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'character', 'shower', 'stereo', 'microphone', 'wall', 'toy', 'cutting_board', 'cup', 'novel', 'food_food', 'check', 'food_onion', 'coffee_pot', 'console', 'shaving_cream', 'towel_rack', 'shoes', 'cd', 'hairbrush', 'maindoor', 'soap', 'coffee']\nTask:Read\nresponse:\n{\n    \"thoughts\": Read a book in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I need to write an e-mail to my boss about the meeting at work tomorrow.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[OPEN] <desk>", "[PUTBACK] <keyboard> <desk>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <mousepad>", "[PUTBACK] <mouse> <mousepad>", "[PLUGIN] <keyboard>", "[PLUGIN] <mouse>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TYPE] <keyboard>", "[GRAB] <mouse>", "[TOUCH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'toilet', 'chair', 'freezer', 'floor', 'ceiling', 'thread', 'table_cloth', 'ceilinglamp', 'faucet', 'food_peanut_butter', 'dining_room', 'wall', 'nightstand', 'powersocket', 'tape', 'photoframe', 'bed', 'food_food', 'cards', 'kitchen_counter', 'dishrack', 'dresser', 'hair', 'pajamas', 'door', 'home_office', 'dishwasher', 'vase']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": go to the kitchen and put the groceries in the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_pizza', 'beer', 'tea', 'microwave', 'toilet', 'chair', 'freezer', 'toothbrush', 'cpuscreen', 'piano_bench', 'faucet', 'mousepad', 'couch', 'nightstand', 'powersocket', 'food_food', 'coffe_maker', 'curtain', 'shaving_cream', 'shoes', 'dishrack', 'dresser', 'door', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room, walk to the phone reach for the phone and pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'beer', 'crayon', 'freezer', 'walllamp', 'bathroom', 'detergent', 'wallshelf', 'floor', 'ceiling', 'purse', 'computer', 'mat', 'tvstand', 'shower', 'window', 'couch', 'food_peanut_butter', 'dining_room', 'wall', 'photoframe', 'bed', 'television', 'curtain', 'diary', 'cupboard', 'controller', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Wipe down baseboards please\nresponse:\n{\n    \"thoughts\": Use a rag to get all the dust off the baseboards please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <wall>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'brush', 'mat', 'tvstand', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'bag', 'light', 'dishwasher', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'food_ice_cream', 'microwave', 'walllamp', 'orchid', 'bills', 'knife', 'wallshelf', 'floor', 'ceiling', 'remote_control', 'computer', 'tvstand', 'faucet', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'light_bulb', 'coffe_maker', 'television', 'curtain', 'lighter', 'vacuum_cleaner', 'dresser', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I grab the remote.  I find the red power circular button. I point the remote at the television.   I push button down and TV screen lights up. I sit and watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'jelly', 'chair', 'blow_dryer', 'bathroom', 'wallshelf', 'shampoo', 'phone', 'desk', 'clothes_jacket', 'stovefan', 'faucet', 'oil', 'closetdrawer', 'food_noodles', 'mousepad', 'microphone', 'towel', 'broom', 'bed', 'newspaper', 'after_shave', 'centerpiece', 'console', 'dresser', 'bathroom_cabinet', 'soap', 'tooth_paste', 'scrabble']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to living room. Find book. Grab book. Open book. Read book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'shelf', 'cup', 'television', 'doorjamb', 'mouse', 'address_book', 'dresser', 'conditioner', 'bathtub', 'spoon', 'light', 'table', 'soap']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on computer. Open the address book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[FIND] <address_book>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'couch', 'microphone', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'photoframe', 'keys', 'television', 'tray', 'doorjamb', 'dresser', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'walllamp', 'orchid', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'powersocket', 'oven', 'bench', 'photoframe', 'television', 'curtain', 'food_onion', 'doorjamb', 'towel_rack', 'bathroom_cabinet', 'light', 'home_office']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'bathroom', 'toothbrush', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'box', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'comb', 'dresser', 'laundry_detergent', 'instrument_guitar', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I reach for the remote. I push the on button on the remote. I sit down on the couch and watch television.", "actions": ["[WALK] <home_office>", "[WALK] <button>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'shoe_rack', 'toilet', 'chair', 'band-aids', 'freezer', 'walllamp', 'cat', 'trashcan', 'folder', 'bookshelf', 'deck_of_cards', 'shower', 'mousepad', 'food_kiwi', 'wall', 'powersocket', 'cup', 'bed', 'clothes_gloves', 'keys', 'television', 'kitchen_counter', 'dishrack', 'controller', 'napkin', 'light', 'soap', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I go to the living room then go to the cat and touch it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'character', 'desk', 'bookshelf', 'mat', 'deck_of_cards', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'bed', 'coffee_filter', 'wine', 'curtain', 'doorjamb', 'kitchen_counter', 'pajamas', 'door', 'food_steak', 'bag', 'light', 'electrical_outlet']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the electric outlet. I switch on the light.", "actions": ["[WALK] <home_office>", "[WALK] <electrical_outlet>", "[FIND] <electrical_outlet>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'bathroom', 'orchid', 'knifeblock', 'food_egg', 'clothes_underwear', 'shampoo', 'bedroom', 'tvstand', 'shoe-shine_kit', 'pillow', 'dining_room', 'bench', 'powersocket', 'coffe_maker', 'video_game_controller', 'television', 'food_rice', 'tray', 'after_shave', 'food_onion', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'comb', 'cupboard', 'bathroom_cabinet', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[TOUCH] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['spectacles', 'microwave', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'wall', 'food_food', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'headset', 'cupboard', 'food_steak', 'light', 'laser_pointer', 'clothes_shirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Find groceries. Grab groceries. Walk to refrigerator. Open refrigerator door. Place groceries into refrigerator. Close refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'chair', 'bathroom', 'detergent', 'trashcan', 'toothbrush', 'knifeblock', 'bathroom_counter', 'character', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'nightstand', 'oven', 'bench', 'powersocket', 'cup', 'bed', 'ice', 'diary', 'cards', 'dresser', 'hairbrush', 'home_office', 'dishwasher', 'scrabble']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": go to radio. run on radio.,\n    \"actions\": ['[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'photoframe', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "go to home office,switch on computer and send the mail by typing with keyboard", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'food_pizza', 'filing_cabinet', 'tablelamp', 'freezer', 'nail_polish', 'cat', 'food_carrot', 'drawing', 'orchid', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'stamp', 'ceilinglamp', 'plate', 'candle', 'dining_room', 'wall', 'bench', 'food_vegetable', 'newspaper', 'food_food', 'television', 'after_shave', 'cards', 'cd', 'clothes_dress', 'table']\nTask:Fix snack\nresponse:\n{\n    \"thoughts\": Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTBACK] <food_carrot> <plate>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[PUTBACK] <food_vegetable> <plate>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'bathroom', 'toothbrush', 'clothes_underwear', 'character', 'stamp', 'clothes_jacket', 'table_cloth', 'stovefan', 'oil', 'clothes_socks', 'clothes_pants', 'towel', 'wall', 'powersocket', 'basket_for_clothes', 'photoframe', 'coffee_filter', 'wooden_spoon', 'kitchen_counter', 'laundry_detergent', 'conditioner', 'napkin', 'bathroom_cabinet', 'soap', 'tooth_paste', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'band-aids', 'freezer', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'envelope', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'mat', 'electric_shaver', 'mousepad', 'towel', 'wall', 'powersocket', 'light_bulb', 'novel', 'curtain', 'doorjamb', 'mouse', 'door', 'light', 'colander', 'board_game']\nTask:Read to child\nresponse:\n{\n    \"thoughts\": Read a book out loud to a child,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['beer', 'microwave', 'cd_player', 'walllamp', 'bathroom', 'detergent', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'computer', 'mat', 'tvstand', 'shower', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'clothes_gloves', 'television', 'doorjamb', 'cd', 'hair', 'controller', 'instrument_guitar', 'clothes_skirt', 'instrument_violin']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to the living room, switch on the CD player (then choose the music cd and insert into the player) and sit on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'crayon', 'chair', 'food_butter', 'food_carrot', 'ceilingfan', 'drawing', 'knife', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'thread', 'brush', 'plate', 'wall', 'bench', 'powersocket', 'oven', 'check', 'food_food', 'doorjamb', 'console', 'dresser', 'cupboard', 'door', 'maindoor', 'table', 'light']\nTask:Eat\nresponse:\n{\n    \"thoughts\": I get my prepared dinner set on the dining table, sit down on chair, pick up utensils, eat food,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[PUTBACK] <food_food> <table>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <knife>', '[GRAB] <knife>', '[FIND] <plate>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'chair', 'detergent', 'orchid', 'drawing', 'floor', 'ceiling', 'desk', 'mail', 'toaster', 'ceilinglamp', 'plate', 'mat', 'shower', 'homework', 'stovefan', 'dish_soap', 'wall', 'wall_clock', 'oven', 'check', 'coffe_maker', 'tray', 'console', 'towel_rack', 'kitchen_counter', 'cupboard', 'controller', 'table', 'dishwasher', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Put the dishes on dinner table in the dishwasher and wash them", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[WALK] <table>", "[FIND] <pot>", "[GRAB] <pot>", "[WALK] <dishwasher>", "[PUTBACK] <pot> <dishwasher>", "[CLOSE] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'freezer', 'food_carrot', 'milk', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'mousepad', 'garbage_can', 'clothes_hat', 'food_orange', 'food_dessert', 'wall', 'powersocket', 'food_vegetable', 'food_cheese', 'doorjamb', 'lighter', 'mouse', 'coffee', 'door', 'light', 'colander', 'teeth']\nTask:Rotate stock in refrigerator\nresponse:\n{\n    \"thoughts\": Throw out expired items, put older items in front.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[TURNTO] <food_bread>', '[LOOKAT] <food_bread>', '[PUTOBJBACK] <food_bread>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[FIND] <garbage_can>', '[PUTBACK] <food_cheese> <garbage_can>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[PUTOBJBACK] <food_cheese>', '[FIND] <food_dessert>', '[TURNTO] <food_dessert>', '[LOOKAT] <food_dessert>', '[PUSH] <food_dessert>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <milk>', '[GRAB] <milk>', '[TURNTO] <milk>', '[LOOKAT] <milk>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'food_butter', 'drawing', 'cpuscreen', 'hanger', 'floor', 'phone', 'desk', 'stamp', 'keyboard', 'computer', 'alcohol', 'toaster', 'laptop', 'mat', 'mousepad', 'wall', 'oven', 'bed', 'mouthwash', 'doorjamb', 'mouse', 'chessboard', 'door', 'clothes_scarf']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "walk to home office and switch on laptop and with the useof mouse and keyboard send mail", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'keyboard', 'computer', 'stovefan', 'brush', 'window', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'oven', 'coffe_maker', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'shaving_cream', 'cards', 'door', 'table', 'light', 'tooth_paste', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": The cat is laying on the couch in the living room. I want to pet it, so I am going to walk to the living room, sit on the couch, and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_bread', 'filing_cabinet', 'tablelamp', 'hanger', 'ceiling', 'computer', 'bookmark', 'stovefan', 'mat', 'plate', 'shower', 'dish_soap', 'mousepad', 'closetdrawer', 'candle', 'food_orange', 'food_cheese', 'photoframe', 'food_food', 'food_rice', 'water', 'tray', 'doorjamb', 'towel_rack', 'headset', 'table', 'light', 'tooth_paste']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen, fins soap, arrange dishes, wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[FIND] <dish_soap>", "[FIND] <water>", "[GRAB] <plate>", "[GRAB] <dish_soap>", "[WASH] <plate>", "[RINSE] <plate>", "[DROP] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'beer', 'chair', 'freezer', 'knifeblock', 'bathroom_counter', 'longboard', 'phone', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'balanceball', 'stereo', 'mousepad', 'clothes_pants', 'nightstand', 'bench', 'bed', 'photoframe', 'novel', 'television', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'food_steak', 'sink', 'hairdryer', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the book shelf, get a book, open the book and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'tablelamp', 'microwave', 'dog', 'milk', 'orchid', 'floor', 'ceiling', 'computer', 'toaster', 'plate', 'shower', 'dish_soap', 'wall', 'food_cheese', 'coffe_maker', 'check', 'coffee_filter', 'oven_mitts', 'cards', 'towel_rack', 'pencil', 'kitchen_counter', 'comb', 'cupboard', 'conditioner', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the  kitchen. Open the dishwasher. Put plates into the dishwasher. Put dish soap into the dishwasher. Close the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'water_glass', 'pot', 'spectacles', 'walllamp', 'food_carrot', 'kitchen_cabinet', 'shampoo', 'longboard', 'computer', 'bookmark', 'mat', 'stereo', 'shower', 'homework', 'oil', 'clothes_socks', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'water', 'food_onion', 'headset', 'cd', 'dresser', 'controller', 'sink', 'napkin', 'clothes_skirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Go to kitchen. Open the the cabinet. Locate the water glass. Pour water into glass. Drag the bottle close to mouth. Drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[CLOSE] <kitchen_cabinet>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'milk', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'button', 'toaster', 'shower', 'oil', 'closetdrawer', 'clothes_hat', 'pillow', 'wall', 'toy', 'box', 'coffe_maker', 'television', 'toothbrush_holder', 'razor', 'wine', 'food_sugar', 'kitchen_counter', 'cupboard', 'pajamas', 'sink', 'coffee_cup', 'dishwasher', 'shoe_rack']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[LOOKAT] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <milk>", "[GRAB] <milk>", "[PUTBACK] <milk> <coffe_maker>", "[FIND] <food_sugar>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[SWITCHON] <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to my home office. I find the computer along with mechanical keyboard on the desk. I sit on chair. I switch on the computer. I type on keyboard and work on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'mail', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to home office and open door.  Walk to desk and sit in computer chair.  Look at computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TOUCH] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'knife', 'cpuscreen', 'bathroom_counter', 'floor', 'keyboard', 'face', 'ceilinglamp', 'mat', 'faucet', 'shower', 'mousepad', 'closetdrawer', 'towel', 'bench', 'oven', 'bed', 'coffe_maker', 'television', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'chessboard', 'hair', 'controller', 'bathtub', 'bathroom_cabinet']\nTask:Rain welcome\nresponse:\n{\n    \"thoughts\": The robot would greet me and then hand me a towel.  It would make an appropriate quip about the weather while I dried off.  It would receive the wet towel when I was done.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <towel>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <face>', '[WIPE] <face>', '[FIND] <hair>', '[WIPE] <hair>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'coin', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch the television", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[SWITCHON] <television>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'trashcan', 'hanger', 'floor', 'bedroom', 'computer', 'button', 'brush', 'mat', 'shower', 'faucet', 'window', 'clothes_hat', 'wall', 'child', 'oven', 'bed', 'newspaper', 'novel', 'curtain', 'centerpiece', 'doorjamb', 'mouse', 'cd', 'dresser', 'hairdryer', 'clothes_scarf', 'vase', 'board_game']\nTask:Story reading time\nresponse:\n{\n    \"thoughts\": Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[TURNTO] <child>', '[LOOKAT] <child>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'kettle', 'mouse', 'dresser', 'hair', 'door', 'coffee']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "walk to living room. look at computer. switch on computer. sit in chair. watch computer. switch off computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[WATCH] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'mat', 'shower', 'mousepad', 'wall', 'powersocket', 'oven', 'bench', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'door', 'table', 'light', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go in my office, sit at my desk turn on the computer, then type on my keyboard and open my web browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'food_cake', 'computer', 'toaster', 'mat', 'pillow', 'wall', 'powersocket', 'cup', 'coffe_maker', 'coffee_filter', 'water', 'console', 'kitchen_counter', 'hairbrush', 'cupboard', 'conditioner', 'hairdryer', 'sink', 'instrument_violin', 'door', 'bathroom_cabinet', 'dishwasher', 'board_game', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I find the coffee machine, fill it with water and coffee then press the button to make the coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'microwave', 'toilet', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'food_egg', 'cpuscreen', 'clothes_underwear', 'tea_bag', 'keyboard', 'faucet', 'clothes_socks', 'food_noodles', 'dining_room', 'bench', 'cup', 'bed', 'check', 'toilet_paper', 'television', 'after_shave', 'console', 'comb', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'board_game', 'teeth']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my cat, and I pet him,\n    \"actions\": ['[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'shaving_cream', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "switch on the computer,  grab a chair and sit in front of the computer. type with keyboard and browse the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'stereo', 'tvstand', 'mousepad', 'couch', 'broom', 'wall', 'bed', 'check', 'toilet_paper', 'television', 'ice', 'kettle', 'mouse', 'door', 'sink', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go in my office, sit at my desk turn on the computer, then type on my keyboard and open my web browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['glue', 'food_salt', 'microwave', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'stovefan', 'ceilinglamp', 'tvstand', 'stereo', 'closetdrawer', 'food_orange', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'television', 'coffee_filter', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'pajamas', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the kitchen, I find the light switch, and then i switch on the lights in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'drawing', 'floor', 'purse', 'computer', 'plate', 'balanceball', 'shower', 'stereo', 'dustpan', 'clothes_hat', 'microphone', 'bench', 'basket_for_clothes', 'cup', 'bed', 'check', 'coffe_maker', 'food_onion', 'after_shave', 'cards', 'pencil', 'chessboard', 'maindoor', 'napkin', 'table', 'board_game', 'clothes_shirt', 'coffee']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'bookshelf', 'keyboard', 'food_cake', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'razor', 'television', 'wine', 'tray', 'doorjamb', 'dresser', 'hair', 'controller', 'door', 'clothes_scarf', 'light', 'instrument_violin']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <bathroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'mop', 'measuring_cup', 'microwave', 'chair', 'freezer', 'ground_coffee', 'bathroom', 'floor', 'ceiling', 'dvd_player', 'bookshelf', 'toaster', 'stovefan', 'faucet', 'iron', 'wall', 'wall_clock', 'cutting_board', 'shelf', 'clothes_gloves', 'coffe_maker', 'coffee_filter', 'food_rice', 'curtain', 'kitchen_counter', 'controller', 'sink', 'clothes_shirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": You walk into the kitchen and turn to the coffee maker. You grab the coffee filter and put into the coffee machine, you also grab the ground coffee and put it into the coffee machine. You turn on the coffee machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'folder', 'desk', 'character', 'tea_bag', 'remote_control', 'purse', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'television', 'doorjamb', 'dresser', 'hairbrush', 'spoon', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'cat', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'ceilinglamp', 'homework', 'mousepad', 'wall', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'shoes', 'comb', 'dresser', 'hairdryer']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I open my laptop and then login to Windows. Then I open Google Chrome, and type gmail.com into the address bar. My email automatically logs in, then I click the Compose button, type in the address of the recipient, then the subject of the email, and then start typing my email,\n    \"actions\": ['[WALK] <laptop>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'beer', 'chair', 'walllamp', 'bathroom', 'cat', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'tooth_paste', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'coin', 'doorjamb', 'dresser', 'board_game']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch TV in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'trashcan', 'bills', 'knife', 'floor', 'desk', 'mail', 'button', 'toaster', 'plate', 'faucet', 'shower', 'foundation', 'couch', 'bowl', 'wall', 'nightstand', 'oven', 'bed', 'clothes_gloves', 'novel', 'coffe_maker', 'check', 'tray', 'sink', 'table', 'dishwasher', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['toilet', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'table_cloth', 'bookmark', 'mat', 'shower', 'faucet', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'oven', 'keys', 'coffe_maker', 'television', 'tray', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'hairdryer', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I open the room door and press the light button on", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_bread', 'microwave', 'freezer', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'box', 'oven_mitts', 'food_onion', 'doorjamb', 'kitchen_counter', 'chessboard', 'dresser', 'cupboard', 'door', 'light', 'soap']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Turning on lights,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <bathroom>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'food_ice_cream', 'chair', 'freezer', 'detergent', 'cpuscreen', 'hanger', 'bookshelf', 'bookmark', 'stovefan', 'mat', 'faucet', 'stereo', 'couch', 'candle', 'dining_room', 'nightstand', 'novel', 'coffe_maker', 'ice', 'doorjamb', 'shaving_cream', 'cd', 'laundry_detergent', 'clothes_dress', 'hairdryer', 'light', 'table', 'colander']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "turn radio on, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'shower', 'mousepad', 'pillow', 'dining_room', 'wall', 'wall_clock', 'nightstand', 'tape', 'bed', 'console', 'mouse', 'clothes_scarf', 'light', 'soap', 'teeth']\nTask:Check email\nresponse:\n{\n    \"thoughts\": Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[WALK] <chair>', '[SIT] <chair>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'tablelamp', 'food_salt', 'microwave', 'chair', 'freezer', 'ground_coffee', 'dough', 'drawing', 'orchid', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'toaster', 'mat', 'ceilinglamp', 'stovefan', 'faucet', 'shower', 'oil', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'powersocket', 'food_food', 'coffe_maker', 'television', 'wine', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'food_steak', 'door', 'sink', 'table', 'dishwasher', 'light']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Open fridge. Take food from fridge. Close fridge. Put food on kitchen counter. Prepare food. Turn on stove. Open kitchen cabinet. Take pot. Close cabinet. Put food in pot. Put pot on stove. Cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <kitchen_counter>", "[PUTBACK] <food_food> <kitchen_counter>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'band-aids', 'freezer', 'orchid', 'cpuscreen', 'floor', 'bookshelf', 'envelope', 'toaster', 'stovefan', 'plate', 'shower', 'faucet', 'foundation', 'oil', 'mousepad', 'food_jam', 'microphone', 'dining_room', 'bench', 'oven', 'toy', 'box', 'food_food', 'food_onion', 'diary', 'kitchen_counter', 'cupboard', 'colander']\nTask:Get out dish\nresponse:\n{\n    \"thoughts\": Open cupboard.  Get out the dish.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[OPEN] <cupboard>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <cupboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'tvstand', 'foundation', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'television', 'food_rice', 'doorjamb', 'shoes', 'dresser', 'door', 'light', 'clothes_skirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to room. I enter room. I find where the lightswitch is. I turn on the lights with the lightswitch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[LOOKAT] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'freezer', 'walllamp', 'detergent', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'stamp', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'couch', 'food_peanut_butter', 'form', 'pillow', 'wall', 'bench', 'novel', 'food_food', 'oven_mitts', 'food_rice', 'diary', 'towel_rack', 'controller', 'maindoor', 'light', 'clothes_skirt']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find groceries. Grab groceries. Find fridge. Walk to fridge. Put groceries in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'orchid', 'bills', 'drawing', 'floor', 'ceiling', 'table_cloth', 'toaster', 'mat', 'faucet', 'mousepad', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'cup', 'keys', 'coffe_maker', 'food_food', 'check', 'water', 'wine', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'light', 'board_game']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to the kitchen and drink some water.", "actions": ["[RUN] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <cup> <sink>", "[FIND] <water>", "[GRAB] <cup>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>", "[DRINK] <cup>", "[PUTBACK] <cup> <sink>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'ceilingfan', 'drawing', 'food_egg', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'television', 'food_rice', 'curtain', 'kitchen_counter', 'hair', 'home_office', 'table', 'light', 'dishwasher', 'dirt']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'crayon', 'microwave', 'walllamp', 'drawing', 'knife', 'wallshelf', 'food_egg', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'mat', 'tvstand', 'stereo', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'oven', 'photoframe', 'television', 'razor', 'oven_mitts', 'doorjamb', 'home_office', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "allow sofa to support weight in order to decrease gravity induced entropy", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'toaster', 'mat', 'tvstand', 'shower', 'shoe-shine_kit', 'window', 'mousepad', 'clothes_socks', 'couch', 'food_noodles', 'pillow', 'wall', 'bed', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'home_office', 'light', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into the bedroom and walk to the light switch. Flip the switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'toilet', 'dog', 'freezer', 'band-aids', 'chair', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'toaster', 'couch', 'wall', 'nightstand', 'cup', 'basket_for_clothes', 'photoframe', 'food_food', 'coffe_maker', 'check', 'lighter', 'pencil', 'dishrack', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'colander']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the radio. Turn on the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'freezer', 'bathroom', 'trashcan', 'orchid', 'cpuscreen', 'hanger', 'character', 'computer', 'mail', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'bed', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'cupboard', 'sink', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the cat. I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'toilet', 'detergent', 'dough', 'bills', 'wallshelf', 'clothes_underwear', 'ceiling', 'keyboard', 'table_cloth', 'faucet', 'electric_shaver', 'mousepad', 'couch', 'dustpan', 'form', 'dining_room', 'newspaper', 'photoframe', 'food_food', 'coffe_maker', 'check', 'wooden_spoon', 'tray', 'rag', 'instrument_guitar', 'soap', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.", "actions": ["[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'tea_bag', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'hair', 'door']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I walk to the Home Office then find the computer and sit in the chair. I switch on the computer and start typing the email. I switch the computer off.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['food_oatmeal', 'microwave', 'freezer', 'bathroom', 'nail_polish', 'dough', 'food_carrot', 'orchid', 'drawing', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'brush', 'mat', 'couch', 'pillow', 'food_cereal', 'wall', 'food_vegetable', 'food_food', 'television', 'console', 'cards', 'kitchen_counter', 'cupboard', 'hair', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_vegetable> <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTIN] <food_carrot> <freezer>", "[FIND] <food_cereal>", "[GRAB] <food_cereal>", "[PUTIN] <food_cereal> <freezer>", "[FIND] <food_oatmeal>", "[GRAB] <food_oatmeal>", "[PUTIN] <food_oatmeal> <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[FIND] <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'door', 'teeth', 'printing_paper']\nTask:Print out papers\nresponse:\n{\n    \"thoughts\": Turn on printer and fill with blank paper. Use computer to click on printing option. Take printed page out of print.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <fax_machine>', '[FIND] <fax_machine>', '[SWITCHON] <fax_machine>', '[FIND] <printing_paper>', '[GRAB] <printing_paper>', '[PUTBACK] <printing_paper> <fax_machine>', '[FIND] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[GRAB] <printing_paper>']\n}\n", "input": "\nScene_Objects:['freezer', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'mat', 'faucet', 'food_noodles', 'couch', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'toy', 'photoframe', 'newspaper', 'food_food', 'coffe_maker', 'television', 'check', 'ice', 'curtain', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'dishwasher', 'teeth']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'tablelamp', 'food_salt', 'microwave', 'chair', 'freezer', 'ground_coffee', 'dough', 'drawing', 'orchid', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'toaster', 'mat', 'ceilinglamp', 'stovefan', 'faucet', 'shower', 'oil', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'powersocket', 'food_food', 'coffe_maker', 'television', 'wine', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'food_steak', 'door', 'sink', 'table', 'dishwasher', 'light']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Walk into kitchen. Open fridge. Take food from fridge. Close fridge. Put food on kitchen counter. Prepare food. Turn on stove. Open kitchen cabinet. Take pot. Close cabinet. Put food in pot. Put pot on stove. Cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <kitchen_counter>', '[PUTBACK] <food_food> <kitchen_counter>', '[CLOSE] <freezer>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[CLOSE] <kitchen_cabinet>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <oven>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'food_salt', 'toilet', 'chair', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'food_bacon', 'tvstand', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'check', 'food_food', 'food_donut', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'napkin', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I enter the bedroom and find the light. Then I turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'food_pizza', 'tablelamp', 'microwave', 'floor', 'ceiling', 'desk', 'bookshelf', 'stovefan', 'mat', 'ceilinglamp', 'laptop', 'window', 'homework', 'oil', 'closetdrawer', 'couch', 'nightstand', 'wall', 'cup', 'bed', 'food_food', 'mouthwash', 'toothbrush_holder', 'oven_mitts', 'razor', 'food_donut', 'check', 'clothes_scarf', 'bathroom_cabinet']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Walk to bedroom. Walk to bed, lie in bed.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'tablelamp', 'chair', 'drawing', 'cpuscreen', 'floor', 'phone', 'character', 'desk', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'tvstand', 'mousepad', 'form', 'pillow', 'wall', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'cards', 'door', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter office. Go to desk. Pick up phone.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'shampoo', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'coin', 'doorjamb', 'mouse', 'sink', 'light', 'soap']\nTask:Write  school paper\nresponse:\n{\n    \"thoughts\": Writing  and drafting a paper for  class work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink']\nTask:Work\nresponse:\n", "output": {"thoughts": "working on computer", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <button>", "[TURNTO] <button>", "[LOOKAT] <button>", "[FIND] <mouse>", "[GRAB] <mouse>", "[GRAB] <button>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'knifeblock', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'television', 'curtain', 'doorjamb', 'cd', 'pajamas', 'door', 'table', 'light', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sitting on the couch watching my favorite television show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'blow_dryer', 'walllamp', 'food_butter', 'dough', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'character', 'tea_bag', 'computer', 'mat', 'tvstand', 'faucet', 'dining_room', 'wall', 'powersocket', 'tape', 'bed', 'towel_rack', 'hairbrush', 'conditioner', 'door', 'sink', 'bag', 'bathroom_cabinet', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'walllamp', 'woman', 'drawing', 'phone', 'desk', 'man', 'toaster', 'laptop', 'mat', 'shower', 'electric_shaver', 'oil', 'form', 'pillow', 'dining_room', 'oven', 'shelf', 'bed', 'standingmirror', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'tooth_paste', 'scrabble']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'floor', 'phone', 'stamp', 'bedroom', 'keyboard', 'table_cloth', 'tvstand', 'closetdrawer', 'food_peanut_butter', 'dustpan', 'clothes_hat', 'bowl', 'nightstand', 'cup', 'food_cheese', 'novel', 'video_game_controller', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'door', 'napkin', 'bathroom_cabinet', 'laser_pointer', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find the cat. Walk over to the cat. Touch the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'oil', 'pillow', 'dining_room', 'wall', 'wall_clock', 'bench', 'standingmirror', 'television', 'curtain', 'doorjamb', 'mouse', 'headset', 'napkin', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <chair>', '[SIT] <chair>', '[TURNTO] <television>', '[WATCH] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['tea', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'deck_of_cards', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'video_game_controller', 'television', 'doorjamb', 'lighter', 'dresser', 'bathroom_cabinet', 'soap', 'scrabble']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "you go into the living room, sit on a chair/sofa, grab remote control, turn on tv, change channels, then watch somethign fun.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'ground_coffee', 'nail_polish', 'orchid', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'food_cake', 'bookshelf', 'mail', 'toaster', 'plate', 'homework', 'pillow', 'dining_room', 'wall', 'bed', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[GRAB] <ground_coffee>', '[PUTOBJBACK] <ground_coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'filing_cabinet', 'chair', 'jelly', 'walllamp', 'cat', 'orchid', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'dvd_player', 'character', 'food_bacon', 'bookmark', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'oil', 'closetdrawer', 'dustpan', 'towel', 'wall', 'check', 'doorjamb', 'cupboard', 'laundry_detergent', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and get the soap and wash my both hand using the soap and rinse my hands and use towel.", "actions": ["[WALK] <dining_room>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[FIND] <towel>", "[RINSE] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'faucet', 'picture', 'window', 'dustpan', 'towel', 'wall', 'nightstand', 'photoframe', 'keys', 'toothbrush_holder', 'food_donut', 'mouse', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Cleaning hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['sponge', 'chair', 'walllamp', 'dough', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shoe-shine_kit', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'novel', 'television', 'doorjamb', 'diary', 'shoes', 'dresser', 'door', 'clothes_shirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room. I sit on the couch. I pick up the remote control. I push the power button.  I push the guide button and look for my favorite show \"The Middle.\"  I click that channel and enjoy my show.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUSH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'dry_pasta', 'crayon', 'band-aids', 'detergent', 'drawing', 'clothes_underwear', 'table_cloth', 'stovefan', 'laptop', 'toaster', 'faucet', 'ceilinglamp', 'closetdrawer', 'wall', 'nightstand', 'oven', 'keys', 'food_food', 'tray', 'coffee_pot', 'pencil', 'headset', 'door', 'hairdryer', 'sink', 'table', 'dishwasher', 'clothes_shirt', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": You walk to the living room and find the cat. You look at the cat and touch it to pet it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['crayon', 'chair', 'walllamp', 'cat', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'button', 'alcohol', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'clothes_hat', 'microphone', 'pillow', 'wall', 'photoframe', 'novel', 'coffe_maker', 'television', 'doorjamb', 'coffee_pot', 'dresser']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the living room and sit on the couch. Turn on television using remote control. Use remote control buttons to navigate to desired channel. Watch television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'microwave', 'freezer', 'walllamp', 'blow_dryer', 'detergent', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'faucet', 'mousepad', 'wall', 'wall_clock', 'oven', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'cards', 'kitchen_counter', 'cupboard', 'sink', 'light', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, grab a cup, fill with water and then drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[FIND] <sink>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'tea_bag', 'computer', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'cup', 'photoframe', 'television', 'after_shave', 'doorjamb', 'kettle', 'dresser', 'pajamas']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room and sit on the sofa to watch tv", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'crayon', 'walllamp', 'trashcan', 'dough', 'floor', 'phone', 'bedroom', 'tea_bag', 'toaster', 'mat', 'plate', 'ceilinglamp', 'dish_soap', 'wall', 'pasta', 'oven', 'cutting_board', 'food_cheese', 'newspaper', 'coffe_maker', 'check', 'food_food', 'tray', 'towel_rack', 'pencil', 'dresser', 'conditioner', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'broom', 'wall', 'nightstand', 'television', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'hairbrush', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Turn on computer, pull up browser, browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'keyboard', 'mat', 'ceilinglamp', 'shower', 'oil', 'food_peanut_butter', 'food_orange', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'pencil', 'dresser', 'door', 'napkin', 'light', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": put away groceries in fridge. open the door, put on shelf, close the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'food_pizza', 'chair', 'ground_coffee', 'orchid', 'needle', 'cpuscreen', 'hanger', 'desk', 'food_cake', 'computer', 'closetdrawer', 'clothes_pants', 'dining_room', 'oven', 'powersocket', 'cup', 'photoframe', 'coffe_maker', 'television', 'coin', 'tray', 'after_shave', 'doorjamb', 'diary', 'towel_rack', 'cupboard', 'door', 'bathroom_cabinet']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the living room, walk to the lamp, switch on the lamp.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'freezer', 'bathroom', 'detergent', 'shampoo', 'desk', 'stamp', 'food_bacon', 'brush', 'ceilinglamp', 'oil', 'soap', 'food_orange', 'towel', 'nightstand', 'video_game_controller', 'razor', 'television', 'wine', 'food_onion', 'mouse', 'shoes', 'hair', 'door', 'sink', 'home_office', 'light', 'vase', 'colander', 'teeth']\nTask:Open door\nresponse:\n{\n    \"thoughts\": I walk up to the door.  I use my hand to turn the door knob to the left.  I pull the door open with the same hand I used to turn the door knob,\n    \"actions\": ['[RUN] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'detergent', 'drawing', 'cpuscreen', 'floor', 'phone', 'character', 'bedroom', 'bookshelf', 'keyboard', 'brush', 'stereo', 'oven', 'cup', 'food_donut', 'curtain', 'kettle', 'towel_rack', 'pencil', 'kitchen_counter', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'home_office', 'dishwasher', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room ,find radio ,switch on radio ,listen to music 4", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'scissors', 'mop', 'microwave', 'toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'tvstand', 'shower', 'stereo', 'mousepad', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'chessboard', 'laundry_detergent', 'door', 'board_game']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_salt', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'computer', 'mat', 'faucet', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'toothbrush_holder', 'wine', 'doorjamb', 'shaving_cream', 'kitchen_counter', 'door', 'sink', 'clothes_scarf', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into living room. Find wall. Find light button. Switch on light button.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'measuring_cup', 'orchid', 'hanger', 'floor', 'ceiling', 'phone', 'tea_bag', 'keyboard', 'purse', 'computer', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'novel', 'food_food', 'television', 'doorjamb', 'coffee_pot', 'vacuum_cleaner', 'dresser', 'hairbrush', 'door', 'napkin', 'home_office', 'board_game']\nTask:Sweep hallway please\nresponse:\n{\n    \"thoughts\": Vacuum all the way up and down the hallway.  Get all the dirt and crumbs out of carpet please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['freezer', 'bathroom', 'food_carrot', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'mat', 'stereo', 'closetdrawer', 'couch', 'food_jam', 'clothes_hat', 'pillow', 'food_kiwi', 'wall', 'food_food', 'television', 'diary', 'mouse', 'pencil', 'kitchen_counter', 'shoes', 'dishrack', 'comb', 'dresser', 'cupboard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I pick up groceries. I open fridge. I put groceries in the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[DROP] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'drawing', 'knife', 'orchid', 'hanger', 'floor', 'keyboard', 'purse', 'bookshelf', 'mail', 'toaster', 'ceilinglamp', 'shower', 'balanceball', 'window', 'couch', 'clothes_pants', 'wall', 'oven', 'powersocket', 'television', 'coin', 'diary', 'cd', 'dresser', 'hairbrush', 'hair', 'conditioner', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'nail_polish', 'cpuscreen', 'hanger', 'ceiling', 'phone', 'bedroom', 'thread', 'mail', 'ceilinglamp', 'foundation', 'food_peanut_butter', 'dustpan', 'dining_room', 'powersocket', 'cup', 'newspaper', 'bed', 'coffe_maker', 'check', 'tray', 'wooden_spoon', 'kitchen_counter', 'bathroom_cabinet', 'table', 'home_office', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'microwave', 'toilet', 'walllamp', 'detergent', 'trashcan', 'floor', 'ceiling', 'envelope', 'toaster', 'plate', 'shower', 'stereo', 'dish_soap', 'wall', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'food_food', 'mouthwash', 'check', 'curtain', 'kitchen_counter', 'cupboard', 'pajamas', 'conditioner', 'sink', 'spoon', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['food_bread', 'tablelamp', 'food_ice_cream', 'freezer', 'bathroom', 'orchid', 'bills', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'stereo', 'closetdrawer', 'nightstand', 'oven', 'photoframe', 'clothes_gloves', 'toilet_paper', 'video_game_controller', 'television', 'doorjamb', 'mouse', 'cupboard', 'controller', 'door', 'spoon', 'bathroom_cabinet', 'light', 'dishwasher', 'instrument_violin']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "The task is browse and surf on the internet regarding a particular topic", "actions": ["[WALK] <bedroom>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[WATCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'detergent', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'laptop', 'shower', 'window', 'oil', 'clothes_socks', 'couch', 'food_orange', 'pillow', 'wall', 'novel', 'coffe_maker', 'television', 'food_rice', 'rag', 'doorjamb', 'cupboard', 'door', 'sink', 'napkin', 'table', 'light']\nTask:Relax\nresponse:\n{\n    \"thoughts\": Lie down on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'nail_polish', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'bedroom', 'tea_bag', 'laptop', 'tvstand', 'clothes_socks', 'closetdrawer', 'couch', 'clothes_hat', 'wall', 'nightstand', 'powersocket', 'photoframe', 'coffe_maker', 'check', 'television', 'doorjamb', 'towel_rack', 'cards', 'dresser', 'door', 'home_office', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the dining room and turn on the lights", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'jelly', 'bathroom', 'knife', 'dvd_player', 'character', 'bedroom', 'thread', 'computer', 'toaster', 'shoe-shine_kit', 'deck_of_cards', 'shower', 'electric_shaver', 'homework', 'food_peanut_butter', 'form', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'novel', 'light_bulb', 'food_food', 'curtain', 'coffee_pot', 'food_steak', 'door', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the radio. Turn on the radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'band-aids', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'iron', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'television', 'doorjamb', 'coffee_pot', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go to home office, turn on the computer, open the internet and browse around the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'water_glass', 'food_salt', 'freezer', 'walllamp', 'blow_dryer', 'detergent', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'stovefan', 'mat', 'iron', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'newspaper', 'keys', 'food_food', 'television', 'kitchen_counter', 'cupboard', 'door', 'bathroom_cabinet', 'tooth_paste']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen ,find fridge ,open fridge ,find water  glass , drink from water glass,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'chair', 'band-aids', 'bathroom_counter', 'character', 'desk', 'keyboard', 'purse', 'mat', 'tvstand', 'faucet', 'stereo', 'shower', 'oven', 'cup', 'clothes_gloves', 'novel', 'television', 'oven_mitts', 'wooden_spoon', 'curtain', 'pencil', 'kitchen_counter', 'paper_towel', 'coffee', 'cupboard', 'laundry_detergent', 'dishwasher', 'hands_both']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walking towards Toilet room, And dispose of human waste. Cleaning your hands with tissue paper or water.", "actions": ["[WALK] <bathroom>", "[WALK] <paper_towel>", "[FIND] <paper_towel>", "[TURNTO] <paper_towel>", "[LOOKAT] <paper_towel>", "[GRAB] <paper_towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'mop', 'chair', 'freezer', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'toaster', 'shower', 'window', 'food_orange', 'wall', 'oven', 'cup', 'photoframe', 'food_food', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'spoon', 'table', 'light', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'headset', 'dresser', 'door', 'sink']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to the home office to find the computer. I turn switch computer on , i open the email. i write the email and  send it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'sponge', 'tablelamp', 'mop', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'stamp', 'wine_glass', 'clothes_socks', 'couch', 'pillow', 'wall', 'nightstand', 'food_cheese', 'bed', 'photoframe', 'food_food', 'check', 'doorjamb', 'door', 'sink', 'bag', 'bathroom_cabinet', 'table', 'light']\nTask:Entertain\nresponse:\n{\n    \"thoughts\": Give guests wine, cheese, crackers and sit and talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <food_cheese>', '[FIND] <food_cheese>', '[FIND] <food_food>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <wine_glass>', '[GRAB] <wine_glass>', '[DRINK] <wine_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'bathroom', 'food_butter', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mop_bucket', 'toaster', 'shower', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'toy', 'bed', 'razor', 'television', 'after_shave', 'doorjamb', 'dresser', 'conditioner', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into the bedroom and walk to the light switch. Flip the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'scissors', 'tea', 'freezer', 'blow_dryer', 'drawing', 'knifeblock', 'cpuscreen', 'clothes_underwear', 'character', 'desk', 'toaster', 'shower', 'picture', 'mousepad', 'candle', 'dining_room', 'bench', 'oven', 'basket_for_clothes', 'bed', 'check', 'television', 'toothbrush_holder', 'curtain', 'mouse', 'sink', 'light', 'tooth_paste', 'laser_pointer']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the living room and then I walk to the light switch in there and turn the lights on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'microwave', 'toilet', 'walllamp', 'bathroom', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'ceilinglamp', 'tvstand', 'foundation', 'mousepad', 'oil', 'couch', 'pillow', 'bowl', 'cutting_board', 'photoframe', 'food_donut', 'tray', 'curtain', 'doorjamb', 'mouse', 'headset', 'hairdryer', 'instrument_guitar', 'colander', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room ,find electrical outlet , switch on eletrical outlet", "actions": ["[WALK] <home_office>", "[WALK] <electrical_outlet>", "[FIND] <electrical_outlet>", "[SWITCHON] <electrical_outlet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'tablelamp', 'microwave', 'chair', 'walllamp', 'floor', 'ceiling', 'character', 'bedroom', 'stovefan', 'ceilinglamp', 'faucet', 'oil', 'closetdrawer', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'nightstand', 'cutting_board', 'photoframe', 'television', 'mouse', 'pencil', 'hairbrush', 'pajamas', 'bathroom_cabinet', 'light', 'teeth']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Bring book, lie on the sofa, read book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[LIE] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'needle', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'computer', 'table_cloth', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'towel', 'wall', 'nightstand', 'oven', 'cup', 'novel', 'television', 'toothbrush_holder', 'mouse', 'pencil', 'door', 'sink', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Reading book , Its free time hobby for relaxing", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[SIT] <chair>", "[FIND] <novel>", "[TOUCH] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tea', 'microwave', 'bathroom', 'trashcan', 'orchid', 'knife', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'envelope', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'food_food', 'food_donut', 'curtain', 'doorjamb', 'kitchen_counter', 'hair', 'door', 'sink', 'table', 'light', 'dishwasher']\nTask:Put out flowers\nresponse:\n{\n    \"thoughts\": Place a flower centrepiece on the entrance hall table.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <centerpiece>', '[FIND] <centerpiece>', '[GRAB] <centerpiece>', '[WALK] <table>', '[PUTBACK] <centerpiece> <table>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'crayon', 'measuring_cup', 'freezer', 'walllamp', 'bathroom', 'knife', 'needle', 'cpuscreen', 'wallshelf', 'orchid', 'ceiling', 'phone', 'character', 'computer', 'bookmark', 'ceilinglamp', 'deck_of_cards', 'picture', 'mousepad', 'closetdrawer', 'couch', 'powersocket', 'oven', 'keys', 'check', 'food_rice', 'dishwasher', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I hear the phone ringing in the living room. I walk to the living room and go to the phone. I pick it up and answer the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>", "[SWITCHON] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'ceilingfan', 'food_carrot', 'toothbrush', 'drawing', 'knife', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'cup', 'photoframe', 'light_bulb', 'coffe_maker', 'television', 'tray', 'curtain', 'pencil', 'maindoor', 'table', 'light', 'coffee']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn TV on to TV channel and sit down on the couch and watch TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[SIT] <couch>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'walllamp', 'dough', 'drawing', 'needle', 'food_egg', 'wallshelf', 'cpuscreen', 'orchid', 'floor', 'ceiling', 'computer', 'mail', 'mat', 'tvstand', 'shoe-shine_kit', 'couch', 'pillow', 'dining_room', 'wall', 'photoframe', 'food_food', 'television', 'coin', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'clothes_dress', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room, turn to the sofa. Sit on the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'jelly', 'ground_coffee', 'knifeblock', 'drawing', 'needle', 'food_egg', 'hanger', 'clothes_underwear', 'desk', 'purse', 'alcohol', 'mat', 'stereo', 'faucet', 'food_peanut_butter', 'wall', 'powersocket', 'photoframe', 'novel', 'coffe_maker', 'curtain', 'mouse', 'vacuum_cleaner', 'shoes', 'bathroom_cabinet', 'clothes_shirt', 'electrical_outlet']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Plug vacuum into the wall. Turn on the vacuum. Move vacuum around the floor to clean it. Unplug vacuum.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <vacuum_cleaner>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[FIND] <electrical_outlet>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[PUTOBJBACK] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'bathroom', 'detergent', 'drawing', 'hanger', 'floor', 'ceiling', 'tea_bag', 'bookshelf', 'piano_bench', 'ceilinglamp', 'closetdrawer', 'candle', 'dustpan', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'bed', 'novel', 'check', 'doorjamb', 'lighter', 'dresser', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the bedroom and use the button on the wall, next to the door to turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'love_seat', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'clothes_scarf']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit and watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'food_pizza', 'scissors', 'tablelamp', 'microwave', 'band-aids', 'floor', 'folder', 'food_cake', 'mail', 'computer', 'table_cloth', 'toaster', 'brush', 'plate', 'shower', 'dish_soap', 'wall', 'coffe_maker', 'oven_mitts', 'food_donut', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'dishrack', 'cupboard', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'measuring_cup', 'toilet', 'chair', 'drawing', 'food_egg', 'clothes_underwear', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'bookshelf', 'button', 'shoe-shine_kit', 'shower', 'window', 'broom', 'wall', 'wall_clock', 'powersocket', 'bench', 'standingmirror', 'novel', 'coin', 'doorjamb', 'door', 'light', 'table', 'creditcard']\nTask:Read\nresponse:\n{\n    \"thoughts\": We pick out a book from the bookshelf and read it.,\n    \"actions\": ['[WALK] <novel>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'freezer', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'computer', 'mop_bucket', 'mat', 'tvstand', 'stereo', 'ceilinglamp', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'keys', 'television', 'ice', 'doorjamb', 'pencil', 'coffee', 'home_office', 'light', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "sitting down to listen to music on the radio", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'microwave', 'toilet', 'freezer', 'trashcan', 'food_carrot', 'milk', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'computer', 'mail', 'mat', 'shower', 'pillow', 'dining_room', 'wall', 'nightstand', 'doorjamb', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'table', 'tooth_paste']\nTask:Get glass of milk\nresponse:\n{\n    \"thoughts\": Get an empty glass. Take milk from refrigerator and open it. Pour milk into glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[POUR] <milk> <cup>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'plate', 'faucet', 'couch', 'pillow', 'dining_room', 'broom', 'wall', 'powersocket', 'box', 'photoframe', 'novel', 'food_food', 'television', 'food_donut', 'curtain', 'doorjamb', 'bathroom_cabinet']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on couch. Grab book. Open book. Read book.", "actions": ["[WALK] <couch>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['band-aids', 'walllamp', 'floor', 'keyboard', 'toaster', 'mat', 'plate', 'tvstand', 'stereo', 'window', 'dish_soap', 'couch', 'food_kiwi', 'wall', 'oven', 'bench', 'cup', 'coffe_maker', 'wine', 'tray', 'cards', 'dishrack', 'dresser', 'clothes_dress', 'maindoor', 'spoon', 'napkin', 'dishwasher', 'vase', 'colander']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'bathroom', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'mousepad', 'dining_room', 'wall', 'light_bulb', 'check', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'clothes_dress', 'door', 'table', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'scissors', 'microwave', 'ceilingfan', 'drawing', 'knifeblock', 'orchid', 'floor', 'ceiling', 'desk', 'bookshelf', 'purse', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'nightstand', 'child', 'cup', 'television', 'curtain', 'centerpiece', 'food_steak', 'hairdryer', 'table', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on tv.  Sit with kids.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[FIND] <child>', '[GREET] <child>', '[FIND] <child>', '[GREET] <child>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'nail_polish', 'drawing', 'orchid', 'wallshelf', 'food_egg', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'bed', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'laser_pointer']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the living room, sit on the couch, find the remote, switch on the TV and watch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'needle', 'floor', 'ceiling', 'phone', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'stereo', 'iron', 'stovefan', 'couch', 'pillow', 'wall', 'television', 'food_rice', 'curtain', 'doorjamb', 'bag', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'crayon', 'jelly', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'dvd_player', 'character', 'keyboard', 'thread', 'food_bacon', 'toaster', 'mat', 'shower', 'closetdrawer', 'pillow', 'dining_room', 'cup', 'newspaper', 'keys', 'mouthwash', 'television', 'colander', 'mouse', 'dishrack', 'sink', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'bathroom', 'bills', 'orchid', 'cpuscreen', 'floor', 'phone', 'character', 'bedroom', 'piano_bench', 'toaster', 'mat', 'ceilinglamp', 'window', 'foundation', 'electric_shaver', 'oil', 'closetdrawer', 'couch', 'broom', 'nightstand', 'wall', 'cup', 'bed', 'check', 'toothbrush_holder', 'sink', 'napkin', 'dishwasher']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <sheets>', '[PULL] <sheets>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'shoe-shine_kit', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'coffe_maker', 'television', 'coin', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Walk to the home office. Walk to the desk. Sit down in chair. Use mouse to open email program. Type email with keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'jelly', 'drawing', 'orchid', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'computer', 'bookmark', 'faucet', 'clothes_socks', 'couch', 'pillow', 'broom', 'wall', 'wall_clock', 'tape', 'photoframe', 'novel', 'check', 'food_onion', 'doorjamb', 'door', 'bathtub', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room and relax on the sofa with a book,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[DROP] <bookmark>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'juice', 'microwave', 'chair', 'freezer', 'detergent', 'food_butter', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'table_cloth', 'toaster', 'ceilinglamp', 'stovefan', 'mat', 'faucet', 'shower', 'couch', 'food_orange', 'pillow', 'wall', 'powersocket', 'oven', 'coffe_maker', 'food_food', 'television', 'food_donut', 'tray', 'doorjamb', 'diary', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'spoon', 'food_steak', 'table', 'dishwasher', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen counter, grab a grocery, walk to fridge, open fridge, put the grocery into fridge, and close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'crayon', 'tea', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'faucet', 'picture', 'wall', 'oven', 'food_food', 'check', 'television', 'ice', 'tray', 'rag', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'spoon']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Take out a rag. Turn on faucet and get water on the rag. Clean counter off.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <rag> <sink>', '[SWITCHOFF] <faucet>', '[GRAB] <rag>', '[SQUEEZE] <rag>', '[WALK] <kitchen_counter>', '[WIPE] <kitchen_counter>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'tea', 'microwave', 'dog', 'freezer', 'walllamp', 'drawing', 'orchid', 'hanger', 'keyboard', 'button', 'ceilinglamp', 'stereo', 'homework', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'food_jam', 'wall', 'oven', 'food_cheese', 'keys', 'tray', 'lighter', 'mouse', 'kitchen_counter', 'cd', 'clothes_dress', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'freezer', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'television', 'ice', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'headset', 'cupboard', 'sink', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'microwave', 'ground_coffee', 'walllamp', 'milk', 'knife', 'needle', 'wallshelf', 'hanger', 'bookshelf', 'clothes_jacket', 'bookmark', 'stovefan', 'mat', 'stereo', 'faucet', 'floor_lamp', 'oil', 'pillow', 'cup', 'curtain', 'after_shave', 'kitchen_counter', 'conditioner', 'sink', 'home_office', 'table', 'dishwasher', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch, flip the switch to the on position", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'cat', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'food_bacon', 'stovefan', 'ceilinglamp', 'iron', 'dustpan', 'pillow', 'wall', 'toy', 'box', 'centerpiece', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cd', 'dresser', 'cupboard', 'light', 'table']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Plug in the vacuum. Turn it on. Push it back and forth across the floor until all the debris is gone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['blender', 'toilet', 'freezer', 'walllamp', 'toothbrush', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'purse', 'computer', 'mat', 'tvstand', 'stereo', 'couch', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'food_food', 'television', 'coffee_filter', 'doorjamb', 'cards', 'pencil', 'table', 'board_game', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I'm tired and want to relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'beer', 'crayon', 'walllamp', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'bookshelf', 'mat', 'iron', 'electric_shaver', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'toy', 'cup', 'light_bulb', 'novel', 'food_food', 'check', 'toothbrush_holder', 'television', 'curtain', 'doorjamb', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the book case, I select a book, I go back and sit down, then I read.,\n    \"actions\": ['[WALK] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['water_glass', 'scissors', 'measuring_cup', 'freezer', 'food_butter', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'food_cake', 'mat', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'food_food', 'check', 'television', 'water', 'kitchen_counter', 'cupboard', 'door', 'light', 'clothes_skirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "go to kitchen, find water glass and grab it, find fridge and open grab water and pour into it, then drink.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[FIND] <freezer>", "[OPEN] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[CLOSE] <freezer>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'walllamp', 'blow_dryer', 'detergent', 'orchid', 'bathroom_counter', 'ceiling', 'phone', 'computer', 'clothes_jacket', 'table_cloth', 'stovefan', 'faucet', 'shower', 'couch', 'clothes_pants', 'wall', 'wall_clock', 'oven', 'bench', 'bed', 'toothbrush_holder', 'tray', 'curtain', 'dresser', 'clothes_dress', 'sink', 'spoon', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open cabinet and collect clothes and put them into washing machine and wash clothes,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_jacket>', '[GRAB] <clothes_jacket>', '[PUTBACK] <clothes_jacket> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'bed', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go the living room,  sit in the chair. Take the remote control and switch on the TV.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'toilet', 'detergent', 'orchid', 'longboard', 'floor', 'desk', 'food_bacon', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'balanceball', 'window', 'bed', 'oven_mitts', 'ice', 'curtain', 'after_shave', 'lighter', 'towel_rack', 'cards', 'controller', 'laundry_detergent', 'maindoor', 'door', 'sink', 'home_office', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Twist the lamp switch until the light goes on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'water_glass', 'tablelamp', 'microwave', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'tea_bag', 'toaster', 'stovefan', 'mat', 'shower', 'faucet', 'ceilinglamp', 'oil', 'wall', 'oven', 'powersocket', 'coffe_maker', 'food_food', 'water', 'ice', 'tray', 'doorjamb', 'towel_rack', 'coffee_pot', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bag', 'table', 'dishwasher', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Going to the kitchen to get a drink of water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[FIND] <sink>", "[PUTBACK] <water_glass> <sink>", "[GRAB] <water_glass>", "[SWITCHOFF] <faucet>", "[DRINK] <water_glass>", "[FIND] <kitchen_counter>", "[PUTBACK] <water_glass> <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'mop', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'candle', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'check', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'door', 'bathroom_cabinet', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I go to the living room, switch off the light, sit in the sofa and do nothing.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHOFF] <light>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'food_jam', 'pillow', 'wall', 'photoframe', 'novel', 'television', 'after_shave', 'doorjamb', 'chessboard', 'dresser', 'laundry_detergent', 'instrument_guitar']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I go into the living room. I find the remote. I turn on the tv. I sit on the couch and watch tv.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ceilingfan', 'toothbrush', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'longboard', 'mop_bucket', 'mat', 'tvstand', 'faucet', 'couch', 'clothes_pants', 'towel', 'wall', 'basket_for_clothes', 'photoframe', 'light_bulb', 'wooden_spoon', 'after_shave', 'coffee', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'laser_pointer', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Turn faucet clockwise to turn on water.  Place hands under stream of water, pick up soap, rub hands with soap.  Put soap back down.  Rub hands together with soap on them.  Rinse hands under water from faucet.  Turn faucet counterclockwise to turn water off.  Dry hands by rubbing them on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[WASH] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'pot', 'toilet', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'kettle', 'shaving_cream', 'kitchen_counter', 'chessboard', 'dresser', 'controller', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I get home from work and walk into the living room. I see the couch and go to sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'stovefan', 'toaster', 'laptop', 'faucet', 'closetdrawer', 'couch', 'bowl', 'dining_room', 'wall', 'oven', 'bench', 'bed', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I switch on the stove and cook my groceries there.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <oven>', '[SWITCHON] <oven>', '[FIND] <bowl>', '[GRAB] <bowl>', '[PUTBACK] <bowl> <oven>', '[FIND] <food_food>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'chair', 'freezer', 'walllamp', 'blow_dryer', 'nail_polish', 'detergent', 'wallshelf', 'shampoo', 'ceiling', 'remote_control', 'envelope', 'computer', 'food_bacon', 'toaster', 'faucet', 'oil', 'dining_room', 'broom', 'cup', 'food_food', 'tray', 'diary', 'towel_rack', 'kitchen_counter', 'spoon', 'table', 'colander']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the TV remote. I change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'freezer', 'ceilingfan', 'food_carrot', 'knifeblock', 'clothes_underwear', 'floor', 'ceiling', 'bedroom', 'desk', 'stamp', 'alcohol', 'food_bacon', 'table_cloth', 'ceilinglamp', 'mousepad', 'food_peanut_butter', 'pillow', 'wall', 'food_food', 'check', 'television', 'towel_rack', 'kettle', 'kitchen_counter', 'dishrack', 'hair', 'dishwasher', 'vase']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'bathroom', 'drawing', 'needle', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stereo', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'toothbrush_holder', 'wooden_spoon', 'doorjamb', 'shoes', 'dresser', 'door', 'light', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on the light", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'dog', 'walllamp', 'ground_coffee', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'shower', 'faucet', 'clothes_socks', 'towel', 'wall', 'cutting_board', 'shelf', 'basket_for_clothes', 'coffe_maker', 'food_rice', 'water', 'tray', 'doorjamb', 'food_steak', 'home_office', 'vase', 'soap', 'teeth']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>', '[TURNTO] <shower>', '[FIND] <water>', '[RINSE] <water>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <water>', '[FIND] <towel>', '[GRAB] <towel>', '[SCRUB] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'chair', 'walllamp', 'drawing', 'wallshelf', 'clothes_underwear', 'bedroom', 'bookshelf', 'keyboard', 'button', 'table_cloth', 'shoe-shine_kit', 'homework', 'clothes_socks', 'clothes_pants', 'powersocket', 'basket_for_clothes', 'photoframe', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'clothes_dress', 'laundry_detergent', 'clothes_skirt', 'dishwasher', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom, take the dirt clothes and put into the washing machine, then put the soap, set the cleaning program and switch on the machine", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_skirt>", "[GRAB] <clothes_skirt>", "[PUTBACK] <clothes_skirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[CLOSE] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'bathroom', 'detergent', 'food_carrot', 'cpuscreen', 'wallshelf', 'phone', 'bedroom', 'desk', 'remote_control', 'envelope', 'computer', 'mail', 'tvstand', 'shower', 'iron', 'mousepad', 'closetdrawer', 'oven', 'powersocket', 'newspaper', 'novel', 'food_food', 'curtain', 'dresser', 'food_steak', 'sink', 'table', 'shoe_rack']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[PUSH] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'band-aids', 'walllamp', 'food_carrot', 'bills', 'drawing', 'cpuscreen', 'orchid', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'bookshelf', 'mop_bucket', 'stovefan', 'mat', 'shower', 'broom', 'wall', 'cup', 'bed', 'newspaper', 'mouthwash', 'cupboard', 'door', 'sink', 'bathroom_cabinet']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to bathroom, find toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'juice', 'microwave', 'chair', 'freezer', 'detergent', 'drawing', 'orchid', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'shower', 'mat', 'ceilinglamp', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'newspaper', 'coffe_maker', 'television', 'ice', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'cupboard', 'table', 'dishwasher', 'creditcard']\nTask:Grab some juice\nresponse:\n{\n    \"thoughts\": I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'freezer', 'blow_dryer', 'bathroom', 'detergent', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'toaster', 'faucet', 'stereo', 'couch', 'pillow', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'bed', 'photoframe', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Washing Hands", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[TOUCH] <soap>", "[PUTOBJBACK] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'tablelamp', 'crayon', 'microwave', 'freezer', 'bathroom', 'trashcan', 'ceilingfan', 'knifeblock', 'clothes_underwear', 'longboard', 'shampoo', 'dvd_player', 'folder', 'mail', 'bookmark', 'toaster', 'shower', 'homework', 'couch', 'bed', 'clothes_gloves', 'towel_rack', 'mouse', 'pencil', 'cupboard', 'maindoor', 'table', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'food_salt', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'desk', 'alcohol', 'stovefan', 'mat', 'plate', 'shower', 'couch', 'food_orange', 'pillow', 'wall', 'oven', 'food_food', 'television', 'tray', 'towel_rack', 'kitchen_counter', 'sauce_pan', 'cupboard', 'spoon', 'napkin', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <spoon>", "[GRAB] <spoon>", "[FIND] <sauce_pan>", "[PUTBACK] <spoon> <sauce_pan>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <plate>", "[PUTBACK] <pot> <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'ground_coffee', 'walllamp', 'food_butter', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'thread', 'stovefan', 'toaster', 'tvstand', 'faucet', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'coffe_maker', 'food_food', 'oven_mitts', 'coffee_filter', 'water', 'wine', 'wooden_spoon', 'controller', 'bathroom_cabinet', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'beer', 'microwave', 'bathroom', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'toaster', 'brush', 'plate', 'shower', 'clothes_socks', 'dining_room', 'wall', 'powersocket', 'tape', 'box', 'cup', 'coffe_maker', 'food_food', 'doorjamb', 'centerpiece', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'cd_player', 'chair', 'toilet', 'walllamp', 'cat', 'hanger', 'floor', 'phone', 'remote_control', 'purse', 'alcohol', 'tvstand', 'iron', 'picture', 'window', 'wall', 'nightstand', 'powersocket', 'bench', 'cutting_board', 'basket_for_clothes', 'after_shave', 'doorjamb', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'door', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.,\n    \"actions\": ['[WALK] <cd_player>', '[SWITCHON] <cd_player>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'toaster', 'mat', 'plate', 'shower', 'iron', 'dish_soap', 'closetdrawer', 'pillow', 'wall', 'oven', 'basket_for_clothes', 'coffe_maker', 'check', 'tray', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'dishwasher', 'board_game', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>", "[GRAB] <plate>", "[WIPE] <plate>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[PUTOBJBACK] <dishrack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'ground_coffee', 'knifeblock', 'hanger', 'floor', 'ceiling', 'purse', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'wall', 'cup', 'razor', 'coin', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": To go kitchen and find a cup. open the faucet and fill water in the cup and drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <cup>', '[GRAB] <cup>', '[TURNTO] <faucet>', '[POINTAT] <faucet>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'mop', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'candle', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'check', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'door', 'bathroom_cabinet', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I go to the living room, switch off the light, sit in the sofa and do nothing.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHOFF] <light>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'stamp', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'dining_room', 'wall', 'toy', 'bed', 'food_food', 'toilet_paper', 'mouthwash', 'doorjamb', 'pencil', 'kitchen_counter', 'comb', 'dresser', 'laundry_detergent', 'napkin', 'home_office', 'bathroom_cabinet', 'dishwasher', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['glue', 'blender', 'food_salt', 'toilet', 'chair', 'toothbrush', 'drawing', 'orchid', 'bathroom_counter', 'bookshelf', 'purse', 'alcohol', 'clothes_jacket', 'ceilinglamp', 'foundation', 'pillow', 'bowl', 'dining_room', 'bed', 'curtain', 'lighter', 'doorjamb', 'towel_rack', 'console', 'chessboard', 'cd', 'clothes_scarf', 'instrument_guitar', 'light', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find a radio there and switch it on to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'freezer', 'ground_coffee', 'detergent', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'faucet', 'shower', 'balanceball', 'food_noodles', 'candle', 'bowl', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'toothbrush_holder', 'rag', 'doorjamb', 'towel_rack', 'conditioner', 'sink', 'bathroom_cabinet', 'soap', 'laser_pointer']\nTask:Wipe down sink\nresponse:\n{\n    \"thoughts\": Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[WIPE] <sink>', '[PUTOBJBACK] <rag>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'floor', 'character', 'keyboard', 'toaster', 'mat', 'faucet', 'stereo', 'couch', 'towel', 'wall', 'wall_clock', 'nightstand', 'cup', 'coffe_maker', 'kitchen_counter', 'shoes', 'cd', 'hairbrush', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I have had my lunch and walked towards the sink and applied soap to my hands and washed cleanly.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'keyboard', 'thread', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'mousepad', 'couch', 'wall', 'photoframe', 'bed', 'toilet_paper', 'food_onion', 'towel_rack', 'mouse', 'kitchen_counter', 'cd', 'hairbrush', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'doorjamb', 'towel_rack', 'pencil', 'shoes', 'dresser', 'sink', 'instrument_violin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I go to the the living room, sit down on the sofa, and then I switch on the TV with the TV remote", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['trashcan', 'dough', 'drawing', 'needle', 'longboard', 'floor', 'desk', 'bedroom', 'computer', 'toaster', 'ceilinglamp', 'stovefan', 'shower', 'window', 'floor_lamp', 'mousepad', 'wall', 'nightstand', 'bed', 'television', 'coffee_filter', 'curtain', 'rag', 'doorjamb', 'cards', 'mouse', 'pencil', 'maindoor', 'sink', 'door']\nTask:Turn on lights\nresponse:\n{\n    \"thoughts\": Walk to your lamp, switch it on, and lay on your bed,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[TURNTO] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'mail', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "go to the office and use the computer to write and email", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mail>", "[OPEN] <mail>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'walllamp', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'thread', 'bookshelf', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'clothes_socks', 'wall', 'pasta', 'powersocket', 'cutting_board', 'cup', 'clothes_gloves', 'novel', 'television', 'rag', 'doorjamb', 'shoes', 'door', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['tea', 'chair', 'ground_coffee', 'food_butter', 'floor', 'ceiling', 'computer', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'brush', 'stereo', 'mat', 'electric_shaver', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'tape', 'cup', 'coffe_maker', 'coffee_filter', 'water', 'towel_rack', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'crayon', 'tea', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'faucet', 'picture', 'wall', 'oven', 'food_food', 'check', 'television', 'ice', 'tray', 'rag', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'spoon']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Take out a rag. Turn on faucet and get water on the rag. Clean counter off.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <rag> <sink>', '[SWITCHOFF] <faucet>', '[GRAB] <rag>', '[SQUEEZE] <rag>', '[WALK] <kitchen_counter>', '[WIPE] <kitchen_counter>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'bed', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "Type on a computer in the home office.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'pot', 'toilet', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'kettle', 'shaving_cream', 'kitchen_counter', 'chessboard', 'dresser', 'controller', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I get home from work and walk into the living room. I see the couch and go to sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'oven', 'photoframe', 'television', 'doorjamb', 'headset', 'dresser', 'bathroom_cabinet', 'soap']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I find the remote then I push the button to switch it to the tv channel i want to watch", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'remote_control', 'computer', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'clothes_socks', 'closetdrawer', 'couch', 'pillow', 'towel', 'wall', 'powersocket', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch and use the remote control to turn on the TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'chair', 'freezer', 'toothbrush', 'drawing', 'wallshelf', 'floor', 'ceiling', 'food_bacon', 'toaster', 'mat', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'newspaper', 'food_food', 'television', 'kitchen_counter', 'cupboard', 'conditioner', 'food_steak', 'sink', 'napkin', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "In the kitchen where I left the bags of groceries, I grab cheese from the grocery bag, I walk over to the refrigerator, open it up, neatly put cheese in the fridge, Close the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'bathroom', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'remote_control', 'computer', 'clothes_jacket', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'food_noodles', 'couch', 'clothes_pants', 'pillow', 'wall', 'oven', 'check', 'television', 'tray', 'curtain', 'sink', 'table', 'light', 'teeth']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": You walk to the living room and sit on the sofa. You grab the remote control, look at it and point at the television, you watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'walllamp', 'nail_polish', 'detergent', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'couch', 'microphone', 'pillow', 'wall', 'pasta', 'cup', 'photoframe', 'bed', 'novel', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'kitchen_counter', 'home_office']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the living room and find the book you want to read. Take it and then go to the couch. Sit on it, open the book and start reading.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'character', 'desk', 'bookshelf', 'mat', 'deck_of_cards', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'bed', 'coffee_filter', 'wine', 'curtain', 'doorjamb', 'kitchen_counter', 'pajamas', 'door', 'food_steak', 'bag', 'light', 'electrical_outlet']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the electric outlet. I switch on the light.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <electrical_outlet>', '[FIND] <electrical_outlet>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_bread', 'toilet', 'blow_dryer', 'bathroom', 'knife', 'orchid', 'cpuscreen', 'hanger', 'kitchen_cabinet', 'ceiling', 'phone', 'character', 'computer', 'stereo', 'homework', 'mousepad', 'pillow', 'food_kiwi', 'oven', 'cup', 'check', 'television', 'razor', 'water', 'curtain', 'rag', 'diary', 'mouse', 'bathroom_cabinet']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a glass of water to drink", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'scissors', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'faucet', 'mousepad', 'food_orange', 'wall', 'powersocket', 'wooden_spoon', 'after_shave', 'doorjamb', 'shaving_cream', 'mouse', 'door', 'table']\nTask:Computer work\nresponse:\n{\n    \"thoughts\": Find the computer, Turn on the computer, find keyboard, find mouse, go online,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'tvstand', 'shower', 'closetdrawer', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'powersocket', 'cup', 'bed', 'coffe_maker', 'food_food', 'doorjamb', 'towel_rack', 'pencil', 'headset', 'chessboard', 'dresser', 'hairbrush', 'door', 'clothes_scarf', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go into my room, find the light,  then turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'bookshelf', 'ceilinglamp', 'tvstand', 'shower', 'mat', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'oven', 'coffe_maker', 'tray', 'curtain', 'coffee_pot', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'hairdryer']\nTask:Answer door\nresponse:\n{\n    \"thoughts\": Walk up to the door.  Open the door and greet whomever is out side,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'dog', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'alcohol', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'food_jam', 'pillow', 'wall', 'nightstand', 'cutting_board', 'novel', 'television', 'toothbrush_holder', 'mouse', 'door', 'dirt']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I pick up a book. I sit in a chair. I open the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'blow_dryer', 'milk', 'bills', 'orchid', 'floor', 'ceiling', 'character', 'food_cake', 'alcohol', 'stovefan', 'homework', 'closetdrawer', 'pillow', 'broom', 'wall', 'oven', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'kitchen_counter', 'shoes', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'nail_polish', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'doorjamb', 'shaving_cream', 'dresser', 'sink', 'home_office', 'bathroom_cabinet', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the sofa and watch TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'bathroom', 'cat', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'bedroom', 'food_bacon', 'piano_bench', 'bookmark', 'stovefan', 'ceilinglamp', 'laptop', 'shower', 'stereo', 'mat', 'couch', 'nightstand', 'wall', 'photoframe', 'coffee_filter', 'towel_rack', 'dishrack', 'controller', 'table', 'clothes_skirt', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk into room, grab phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'bathroom', 'drawing', 'knife', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'shower', 'oil', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'pasta', 'bed', 'photoframe', 'novel', 'television', 'curtain', 'doorjamb', 'dresser', 'door', 'napkin', 'light', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk towards to the bedroom, and I open the door of the bedroom, and I switch on the lights inside the bedroom, the switch is beside the door.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'dry_pasta', 'ceilingfan', 'trashcan', 'drawing', 'longboard', 'floor', 'ceiling', 'bedroom', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'shower', 'deck_of_cards', 'couch', 'food_kiwi', 'wall', 'tape', 'basket_for_clothes', 'keys', 'coffe_maker', 'toilet_paper', 'television', 'curtain', 'cd', 'controller', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch. Get comfortable. Switch on the television. Stare at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'novel', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I walk to the home office. I sit on the chair. I switch on the computer. I look at my monitor. I open the browsing application by click the application using mouse. I type the book that I need to searching for (using keyboard). I found the book that I need. I read the book. I switch on the printer. I print the information that I get. I read carefully the book information that I got. I switch off the printer. I switch of the computer.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <fax_machine>", "[SWITCHON] <fax_machine>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[SWITCHOFF] <fax_machine>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'toothbrush', 'drawing', 'floor', 'desk', 'tea_bag', 'toaster', 'mat', 'deck_of_cards', 'window', 'oil', 'mousepad', 'couch', 'pillow', 'broom', 'wall', 'bench', 'cutting_board', 'standingmirror', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'lighter', 'maindoor', 'napkin', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Find the sofa and sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'dough', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'stovefan', 'mousepad', 'wall', 'nightstand', 'television', 'curtain', 'after_shave', 'mouse', 'kitchen_counter', 'clothes_dress', 'door', 'sink', 'home_office', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "find a book to read. sit down. read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'toilet', 'chair', 'cat', 'trashcan', 'ceilingfan', 'orchid', 'cpuscreen', 'longboard', 'desk', 'remote_control', 'button', 'food_bacon', 'foundation', 'couch', 'dining_room', 'oven', 'bench', 'basket_for_clothes', 'cup', 'bed', 'check', 'toothbrush_holder', 'doorjamb', 'coffee_pot', 'vacuum_cleaner', 'controller', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk into living room. Find television remote control. Pick up remote control. Press button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'toilet', 'freezer', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'food_jam', 'pillow', 'dining_room', 'wall', 'powersocket', 'photoframe', 'bed', 'ice', 'curtain', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'instrument_guitar', 'light', 'shoe_rack']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the Living Room. Switch on 1 light. Walk to dining room. Switch on 1 light. Walk to Entrance hall. Switch on 1 light.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'food_cake', 'envelope', 'computer', 'mop_bucket', 'laptop', 'mat', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'television', 'doorjamb', 'mouse', 'controller', 'door']\nTask:Surf internet\nresponse:\n{\n    \"thoughts\": SIT ON COUCH OR CHAIR AND SURF SHOPPING AN SOCIAL MEDIA SITES,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <couch>', '[SIT] <couch>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'toilet', 'jelly', 'band-aids', 'dog', 'walllamp', 'bathroom_counter', 'floor', 'bedroom', 'keyboard', 'mop_bucket', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'couch', 'food_peanut_butter', 'dining_room', 'wall', 'pasta', 'powersocket', 'nightstand', 'razor', 'food_donut', 'tray', 'chessboard', 'controller', 'laundry_detergent', 'light']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. I walk over to the toilet. I sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'microwave', 'freezer', 'trashcan', 'bathroom_counter', 'ceiling', 'desk', 'stovefan', 'stereo', 'shower', 'mousepad', 'dustpan', 'clothes_pants', 'dining_room', 'powersocket', 'bench', 'bed', 'toilet_paper', 'food_food', 'wine', 'mouse', 'pencil', 'kitchen_counter', 'cd', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk to living room ,find radio ,switch on radio ,listen to music 4,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'check', 'television', 'doorjamb', 'mouse', 'address_book', 'dresser', 'door']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Walk to Home office. Sit on the chair. Switch on the computer. Find address in addressbook. Type in the keyboard. Write mail.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <address_book>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'toy', 'photoframe', 'bed', 'television', 'doorjamb', 'towel_rack', 'dresser', 'dishwasher', 'shoe_rack']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Sit down on couch in front of TV.  Use remote to turn on the TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'chair', 'toothbrush', 'drawing', 'floor', 'ceiling', 'computer', 'toaster', 'stovefan', 'mat', 'shower', 'faucet', 'couch', 'wall', 'oven', 'cup', 'coffe_maker', 'food_food', 'mouthwash', 'drinking_glass', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'pajamas', 'sink', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'juice', 'microwave', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'envelope', 'mop_bucket', 'toaster', 'mat', 'faucet', 'dish_soap', 'couch', 'wall', 'cup', 'photoframe', 'toilet_paper', 'video_game_controller', 'wooden_spoon', 'tray', 'console', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'light']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[SWITCHOFF] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <cup>', '[GRAB] <cup>', '[SCRUB] <cup>', '[PUTOBJBACK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[SCRUB] <cup>', '[PUTOBJBACK] <cup>', '[PUTBACK] <sponge> <sink>', '[SWITCHON] <faucet>', '[RINSE] <cup>', '[GRAB] <cup>', '[FIND] <dishrack>', '[PUTBACK] <cup> <dishrack>', '[RINSE] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishrack>', '[RINSE] <sponge>', '[SWITCHOFF] <faucet>', '[GRAB] <sponge>', '[PUTOBJBACK] <sponge>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'mop', 'microwave', 'dog', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'form', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'centerpiece', 'sink', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I bend down. I sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'tablelamp', 'crayon', 'freezer', 'knifeblock', 'bathroom_counter', 'shampoo', 'ceiling', 'thread', 'faucet', 'shower', 'mousepad', 'food_kiwi', 'nightstand', 'wall', 'bench', 'powersocket', 'cup', 'clothes_gloves', 'novel', 'food_food', 'coffe_maker', 'toothbrush_holder', 'water', 'doorjamb', 'mouse', 'shoes', 'dresser', 'cupboard', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Drinking a glass of water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'kitchen_cabinet', 'bathroom_counter', 'desk', 'bedroom', 'thread', 'faucet', 'shower', 'electric_shaver', 'microphone', 'towel', 'dining_room', 'wall', 'oven', 'cutting_board', 'photoframe', 'water', 'drinking_glass', 'curtain', 'food_onion', 'doorjamb', 'kettle', 'kitchen_counter', 'dresser', 'laundry_detergent', 'light', 'table', 'dirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "You walk to the kitchen and open the kitchen cabinet and grab a drinking glass. You find the water and pour the water into the drinking glass. You drink the water from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'cat', 'ceilingfan', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'dvd_player', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'foundation', 'oil', 'couch', 'towel', 'wall', 'pasta', 'powersocket', 'television', 'curtain', 'doorjamb', 'shoes', 'pajamas', 'maindoor', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'tea', 'beer', 'toilet', 'chair', 'orchid', 'knife', 'wallshelf', 'bathroom_counter', 'floor', 'bedroom', 'bookshelf', 'alcohol', 'stovefan', 'mat', 'shower', 'clothes_hat', 'pillow', 'wall', 'powersocket', 'bed', 'newspaper', 'novel', 'razor', 'television', 'doorjamb', 'mouse', 'dresser', 'hairdryer']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to bathroom, I open the door of the bathroom, and sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'oil', 'couch', 'wall', 'cup', 'food_food', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'console', 'hairbrush', 'cupboard', 'food_steak', 'maindoor', 'door', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch favorite shows, netflix, hulu,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['walllamp', 'detergent', 'milk', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'stamp', 'food_cake', 'computer', 'piano_bench', 'mat', 'tvstand', 'faucet', 'shower', 'couch', 'clothes_hat', 'pillow', 'wall', 'powersocket', 'toy', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'mouse', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and I lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'cat', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'ceilinglamp', 'homework', 'mousepad', 'wall', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'shoes', 'comb', 'dresser', 'hairdryer']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I open my laptop and then login to Windows. Then I open Google Chrome, and type gmail.com into the address bar. My email automatically logs in, then I click the Compose button, type in the address of the recipient, then the subject of the email, and then start typing my email,\n    \"actions\": ['[WALK] <laptop>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'tablelamp', 'microwave', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_bacon', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'food_food', 'television', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "go to the kitchen and cook some thing to eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <oven>", "[OPEN] <oven>", "[FIND] <food_pizza>", "[GRAB] <food_pizza>", "[PUTIN] <food_pizza> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'newspaper', 'television', 'curtain', 'doorjamb', 'door', 'home_office', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room. look at remote control. grab remote control. plug in television. switch on television. lie in sofa. point at remote control. watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[LIE] <couch>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'television', 'doorjamb', 'mouse', 'dresser', 'hairbrush', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office. Sit in chair. Switch on computer. Look at computer. Type with keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'filing_cabinet', 'tablelamp', 'beer', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'cat', 'cpuscreen', 'phone', 'character', 'food_cake', 'computer', 'piano_bench', 'brush', 'clothes_pants', 'dining_room', 'nightstand', 'cutting_board', 'toilet_paper', 'coffe_maker', 'check', 'curtain', 'kitchen_counter', 'table', 'scrabble', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Every time this animal is seen, you must approach it and show compassion for it by petting it in a gentle manner.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <bedroom>', '[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[FIND] <brush>', '[GRAB] <brush>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'iron', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'food_donut', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "walk into Living room, look for computer, walk to computer, turn on computer, grab mouse, click mouse", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[WALK] <computer>", "[TURNTO] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'dog', 'bathroom', 'food_carrot', 'bills', 'ceiling', 'phone', 'keyboard', 'stovefan', 'stereo', 'window', 'oven', 'bench', 'tape', 'bed', 'light_bulb', 'food_onion', 'coffee_pot', 'kettle', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'maindoor', 'instrument_guitar', 'table', 'bathroom_cabinet', 'dishwasher', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my living room, find my cat, scrub it using my hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'ground_coffee', 'drawing', 'food_egg', 'cpuscreen', 'hanger', 'floor', 'bedroom', 'tea_bag', 'bookshelf', 'mat', 'tvstand', 'faucet', 'wall', 'nightstand', 'powersocket', 'bed', 'television', 'coffee_filter', 'oven_mitts', 'mouse', 'door', 'clothes_scarf', 'spoon', 'napkin', 'light', 'sink', 'bag']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "go t the bedroom and go to sleep", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'picture', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <chair>', '[SIT] <chair>', '[TURNTO] <television>', '[WATCH] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'bills', 'cpuscreen', 'orchid', 'hanger', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'newspaper', 'check', 'television', 'coin', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Work on Computer", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'toilet_paper', 'television', 'doorjamb', 'dresser', 'cupboard', 'light', 'clothes_shirt']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Go to the living room, Switch on the television. Find the remote and press the change channel button, change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[TOUCH] <button>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'mat', 'tvstand', 'faucet', 'stereo', 'picture', 'mousepad', 'couch', 'microphone', 'pillow', 'wall', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'lighter', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to living room and lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'ground_coffee', 'cat', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'food_bacon', 'mat', 'iron', 'couch', 'food_peanut_butter', 'food_orange', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'clothes_gloves', 'food_food', 'television', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'hair', 'clothes_dress', 'light', 'table']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I put my groceries into the fridge.,\n    \"actions\": ['[WALK] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'toilet', 'cat', 'toothbrush', 'floor', 'ceiling', 'desk', 'tea_bag', 'toaster', 'mat', 'shower', 'couch', 'towel', 'box', 'cup', 'photoframe', 'food_food', 'water', 'towel_rack', 'console', 'mouse', 'vacuum_cleaner', 'hairdryer', 'bathroom_cabinet', 'dishwasher', 'soap', 'tooth_paste', 'dirt', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[FIND] <toothbrush>", "[FIND] <water>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'beer', 'crayon', 'freezer', 'walllamp', 'bathroom', 'detergent', 'wallshelf', 'floor', 'ceiling', 'purse', 'computer', 'mat', 'tvstand', 'shower', 'window', 'couch', 'food_peanut_butter', 'dining_room', 'wall', 'photoframe', 'bed', 'television', 'curtain', 'diary', 'cupboard', 'controller', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Wipe down baseboards please\nresponse:\n{\n    \"thoughts\": Use a rag to get all the dust off the baseboards please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <wall>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'toothbrush', 'drawing', 'orchid', 'wallshelf', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'clothes_socks', 'closetdrawer', 'couch', 'pillow', 'wall', 'toy', 'photoframe', 'television', 'doorjamb', 'dresser', 'controller', 'food_steak', 'door']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the remote. Pick up the remote. Find the sofa. Sit on the sofa. Look at the television. Aim the remote at the television. Press the power button. Flip the channels.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[PUSH] <remote_control>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'tvstand', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'cup', 'television', 'doorjamb', 'mouse', 'shoes', 'door', 'sink', 'napkin', 'clothes_scarf', 'table', 'light', 'dishwasher', 'colander']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room ,find sofa ,sit on sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'ground_coffee', 'bathroom', 'shampoo', 'character', 'bedroom', 'bookshelf', 'keyboard', 'envelope', 'bookmark', 'shower', 'iron', 'foundation', 'mousepad', 'closetdrawer', 'couch', 'wall', 'oven', 'basket_for_clothes', 'cup', 'bed', 'food_food', 'curtain', 'towel_rack', 'cupboard', 'door', 'food_steak', 'bathroom_cabinet', 'dishwasher', 'dirt']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "go to the bathroom and brush your teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'detergent', 'trashcan', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'closetdrawer', 'wall', 'oven', 'cup', 'novel', 'water', 'drinking_glass', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'door']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to the kitchen,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <water>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'food_carrot', 'knife', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'stamp', 'purse', 'computer', 'mat', 'shoe-shine_kit', 'shower', 'foundation', 'mousepad', 'closetdrawer', 'food_jam', 'powersocket', 'coffe_maker', 'television', 'tray', 'curtain', 'diary', 'cards', 'pencil', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the radio.  I switch on the radio and listen", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'tablelamp', 'bathroom', 'trashcan', 'food_butter', 'cpuscreen', 'desk', 'bookshelf', 'computer', 'mop_bucket', 'mat', 'closetdrawer', 'form', 'dining_room', 'nightstand', 'bench', 'oven', 'powersocket', 'toy', 'photoframe', 'coffe_maker', 'television', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'sink', 'home_office']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Grab toys off floor.  Put it in a container.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'chair', 'walllamp', 'drawing', 'needle', 'wallshelf', 'bills', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'form', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'check', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'tooth_paste']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I enter the living room and sit on the couch. I grab the remote and turn on the TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'chair', 'ceilingfan', 'dough', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'folder', 'bedroom', 'character', 'ceilinglamp', 'mat', 'balanceball', 'mousepad', 'wall', 'nightstand', 'box', 'check', 'television', 'doorjamb', 'pencil', 'vacuum_cleaner', 'dresser', 'cupboard', 'conditioner', 'door', 'light', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the kitchen, I find the light switch, and then i switch on the lights in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'freezer', 'blow_dryer', 'bathroom', 'cpuscreen', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'food_bacon', 'tvstand', 'clothes_socks', 'oil', 'towel', 'nightstand', 'oven', 'bed', 'check', 'coffee_filter', 'curtain', 'clothes_dress', 'door', 'sink', 'home_office', 'table', 'light', 'colander', 'laser_pointer']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Pour water into a glass and drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'tea_bag', 'ceilinglamp', 'tvstand', 'shower', 'window', 'homework', 'couch', 'dustpan', 'food_orange', 'dining_room', 'wall', 'nightstand', 'photoframe', 'clothes_gloves', 'coffe_maker', 'check', 'television', 'mouse', 'pencil', 'table', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, turn to the sofa. Sit on the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['mop', 'freezer', 'bathroom', 'dough', 'drawing', 'clothes_underwear', 'shampoo', 'bedroom', 'bookshelf', 'stovefan', 'tvstand', 'shower', 'mousepad', 'couch', 'dining_room', 'wall', 'food_food', 'check', 'rag', 'mouse', 'kitchen_counter', 'chessboard', 'cd', 'shoes', 'door', 'napkin', 'home_office', 'dishwasher', 'board_game', 'coffee']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "i go find the phone. i pick up the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'chair', 'walllamp', 'detergent', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'pillow', 'wall', 'powersocket', 'newspaper', 'television', 'tray', 'doorjamb', 'mouse', 'pencil', 'door', 'light', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I walk into my office and go to the computer desk. I sit at my desk and turn the computer on. Then I write an email.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'alcohol', 'stovefan', 'mat', 'shoe-shine_kit', 'tvstand', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'wall', 'powersocket', 'photoframe', 'food_food', 'check', 'television', 'coffe_maker', 'ice', 'kitchen_counter', 'cupboard', 'hair', 'table', 'scrabble']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'faucet', 'mousepad', 'wall', 'nightstand', 'cup', 'coffe_maker', 'doorjamb', 'mouse', 'chessboard', 'controller', 'bathroom_cabinet', 'light']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": enter office, find computer, turn it on,,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TOUCH] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'glue', 'jelly', 'freezer', 'detergent', 'food_carrot', 'orchid', 'clothes_underwear', 'hanger', 'floor', 'bedroom', 'thread', 'bookmark', 'mat', 'stereo', 'shower', 'mousepad', 'oil', 'nightstand', 'light_bulb', 'coffe_maker', 'television', 'tray', 'pencil', 'shoes', 'bathroom_cabinet', 'dishwasher', 'colander', 'dirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Go to the living room.  Find the telephone.  Lift up the receiver.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'bathroom', 'orchid', 'bills', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'bowl', 'nightstand', 'wall', 'oven', 'bed', 'novel', 'food_food', 'television', 'oven_mitts', 'curtain', 'dresser', 'controller', 'table', 'dishwasher', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room ,find television , switch on remote control,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[SWITCHON] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'food_onion', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "turn on the computer and browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'microwave', 'chair', 'ground_coffee', 'walllamp', 'detergent', 'trashcan', 'drawing', 'knife', 'floor', 'ceiling', 'thread', 'toaster', 'dustpan', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'chessboard', 'cupboard', 'spoon', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find coffee maker. Find coffee filter. Put it in the coffee maker. Find ground coffee. Put it in the coffee maker. Find Water. Put it in the coffe maker. Switch on the coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[FIND] <water>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to the living room, where my computer is. I locate it, switch it on and type desired URL adress in the web browser.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'walllamp', 'trashcan', 'knifeblock', 'drawing', 'needle', 'floor', 'ceiling', 'clothes_jacket', 'stovefan', 'toaster', 'window', 'homework', 'oil', 'dustpan', 'bowl', 'wall', 'pasta', 'oven', 'toy', 'food_food', 'oven_mitts', 'tray', 'kitchen_counter', 'napkin', 'table', 'dishwasher', 'laser_pointer', 'teeth']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I switch on the stove and cook my groceries there.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <oven>', '[SWITCHON] <oven>', '[FIND] <bowl>', '[GRAB] <bowl>', '[PUTBACK] <bowl> <oven>', '[FIND] <food_food>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'mop', 'microwave', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'newspaper', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to the Home Office.  Find the computer.  Turn on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'crayon', 'freezer', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'shampoo', 'desk', 'mail', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'stovefan', 'mat', 'electric_shaver', 'mousepad', 'clothes_pants', 'wall', 'nightstand', 'pasta', 'basket_for_clothes', 'food_food', 'diary', 'towel_rack', 'dresser', 'clothes_dress']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": in my home i walk to the kitchen with some groceries and locate the fridge,then open the fridge put the groceries into the fridge and closed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'needle', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to office, sit at desk, turn on computer, enter password, open application and begin work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[PUSH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'computer', 'stovefan', 'faucet', 'window', 'mousepad', 'closetdrawer', 'couch', 'towel', 'wall', 'nightstand', 'curtain', 'towel_rack', 'mouse', 'coffee_pot', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom and locate the sink. I turn the faucet so that water is running, and rinse my hands briefly. Then I grab the soap and squeeze some onto my hands. I spend a few minutes rubbing my hands together, then I rinse them again. Finally, I turn the faucet off and wipe my hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SQUEEZE] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'bills', 'cpuscreen', 'orchid', 'hanger', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go in my office, sit at my desk turn on the computer, then type on my keyboard and open my web browser.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'cutting_board', 'cup', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'controller', 'door', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'detergent', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'purse', 'toaster', 'mat', 'faucet', 'iron', 'mousepad', 'wall', 'powersocket', 'cup', 'coffe_maker', 'water', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'clothes_scarf', 'colander']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <cup>", "[GRAB] <cup>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'walllamp', 'trashcan', 'food_carrot', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'food_chicken', 'desk', 'stovefan', 'ceilinglamp', 'toaster', 'balanceball', 'shower', 'faucet', 'food_noodles', 'bowl', 'wall', 'oven', 'powersocket', 'cup', 'bed', 'food_food', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'keys', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I need to send an email to my friend so I go into my office and sit down at the desk. I turn on the computer and type the email.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[TURNTO] <desk>", "[LOOKAT] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'jelly', 'freezer', 'bathroom', 'detergent', 'food_butter', 'food_carrot', 'cpuscreen', 'bedroom', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'mousepad', 'food_noodles', 'food_orange', 'nightstand', 'cup', 'check', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'napkin', 'spoon', 'table', 'home_office', 'soap']\nTask:Open door\nresponse:\n{\n    \"thoughts\": Open the door when leaving or coming in.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'toilet', 'walllamp', 'cat', 'food_carrot', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'mop_bucket', 'stovefan', 'mat', 'stereo', 'couch', 'wall', 'powersocket', 'bed', 'novel', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'controller', 'door', 'food_steak', 'spoon', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.", "actions": ["[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'window', 'couch', 'pillow', 'wall', 'oven', 'coffe_maker', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'food_steak', 'door', 'sink', 'table']\nTask:Make toast\nresponse:\n{\n    \"thoughts\": Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[CLOSE] <freezer>', '[FIND] <toaster>', '[PLUGIN] <toaster>', '[PUTBACK] <food_bread> <toaster>', '[SWITCHON] <toaster>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'shoe-shine_kit', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light', 'laser_pointer']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "i'm walkin to the home office, i'm walking to the desk, i'm sitting on the chair, i'm turning on the computer and i'm browsing the internet.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'mousepad', 'pillow', 'wall', 'nightstand', 'photoframe', 'razor', 'television', 'food_rice', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathroom_cabinet', 'table', 'dirt']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to my office. I switch the computer on and sit on the office chair. I look at the computer screen to see what needs to be done. I use my mouse to navigate to the relevant directory, then release it. I type code with my keyboard. When I'm done, I look at the computer again to evaluate what I've done. Then I switch off the computer using the mouse and put back the mouse.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[GRAB] <mouse>', '[SWITCHOFF] <computer>', '[PUTOBJBACK] <mouse>']\n}\n", "input": "\nScene_Objects:['scissors', 'tablelamp', 'beer', 'toilet', 'walllamp', 'orchid', 'bathroom_counter', 'dvd_player', 'stamp', 'envelope', 'clothes_jacket', 'toaster', 'ceilinglamp', 'faucet', 'food_noodles', 'pillow', 'bowl', 'nightstand', 'coffe_maker', 'check', 'ice', 'tray', 'after_shave', 'mouse', 'hairbrush', 'sink', 'napkin', 'light', 'clothes_shirt', 'creditcard']\nTask:Receive credit card\nresponse:\n", "output": {"thoughts": "hear the dooble, go to door, take de envelope, open it with a scissor, take the creditcard", "actions": ["[WALK] <home_office>", "[WALK] <envelope>", "[FIND] <envelope>", "[GRAB] <envelope>", "[WALK] <home_office>", "[WALK] <scissors>", "[FIND] <scissors>", "[OPEN] <envelope>", "[FIND] <creditcard>", "[GRAB] <creditcard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'fork', 'beer', 'walllamp', 'bathroom', 'orchid', 'kitchen_cabinet', 'floor', 'phone', 'toaster', 'mat', 'plate', 'tvstand', 'window', 'dish_soap', 'couch', 'bowl', 'wall', 'oven', 'bench', 'coffe_maker', 'tray', 'rag', 'console', 'kitchen_counter', 'dishrack', 'hair', 'maindoor', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>', '[GRAB] <plate>', '[WIPE] <plate>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[PUTOBJBACK] <dishrack>']\n}\n", "input": "\nScene_Objects:['sponge', 'toilet', 'chair', 'walllamp', 'bathroom', 'ground_coffee', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'bookshelf', 'toaster', 'tvstand', 'faucet', 'shower', 'foundation', 'mousepad', 'wall', 'cup', 'newspaper', 'bed', 'coffe_maker', 'doorjamb', 'dresser', 'conditioner', 'door', 'light', 'dishwasher', 'tooth_paste']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "take remote control, push \"change channel\" button", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'tea', 'ground_coffee', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'button', 'mat', 'deck_of_cards', 'plate', 'tvstand', 'closetdrawer', 'couch', 'food_jam', 'dining_room', 'wall', 'powersocket', 'basket_for_clothes', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'chessboard', 'door', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk into living room, walk to light, turn on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'microwave', 'drawing', 'orchid', 'wallshelf', 'floor', 'table_cloth', 'toaster', 'ceilinglamp', 'plate', 'shower', 'faucet', 'dish_soap', 'closetdrawer', 'pillow', 'wall', 'novel', 'coffe_maker', 'food_rice', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'clothes_scarf', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find dishwasher. Open dishwasher. Find dishrack. Find plate 1. Grab plate 1. Put plate 1 in dishrack. Repeat for dish 2. Find dish soap. Grab dish soap. Put dish soap to dishwasher. Close dishwasher. Switch on dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'orchid', 'hanger', 'ceiling', 'keyboard', 'piano_bench', 'tvstand', 'shower', 'electric_shaver', 'clothes_socks', 'couch', 'wall', 'powersocket', 'bench', 'cutting_board', 'toy', 'photoframe', 'bed', 'razor', 'wooden_spoon', 'after_shave', 'coffee_pot', 'dresser', 'clothes_dress', 'bathtub', 'bathroom_cabinet', 'light', 'clothes_shirt', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to kitchen. grab clothes_dress. look at washing machine. open washing machine. put clothes_dress on washing machine. close washing machine. switch on washing machine. switch off washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[TURNTO] <washing_machine>', '[LOOKAT] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>', '[SWITCHOFF] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'jelly', 'freezer', 'food_carrot', 'drawing', 'floor', 'ceiling', 'desk', 'envelope', 'table_cloth', 'mat', 'faucet', 'mousepad', 'couch', 'pillow', 'towel', 'wall', 'cup', 'food_food', 'mouthwash', 'television', 'water', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'food_steak', 'light', 'dishwasher', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[TURNTO] <cup>", "[LOOKAT] <cup>", "[WALK] <cup>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[WALK] <freezer>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'detergent', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'thread', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'photoframe', 'toilet_paper', 'television', 'razor', 'coin', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Push power button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'dog', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'mat', 'faucet', 'shower', 'couch', 'wall', 'powersocket', 'newspaper', 'keys', 'wooden_spoon', 'doorjamb', 'centerpiece', 'kitchen_counter', 'door', 'home_office', 'light', 'shoe_rack']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Switch on light in home office", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'chair', 'knifeblock', 'knife', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'character', 'desk', 'bedroom', 'bookshelf', 'food_bacon', 'mat', 'shoe-shine_kit', 'food_jam', 'wall', 'powersocket', 'oven', 'photoframe', 'novel', 'mouthwash', 'coffee_filter', 'tray', 'curtain', 'vacuum_cleaner', 'dresser', 'bathtub', 'soap']\nTask:Read to child\nresponse:\n{\n    \"thoughts\": Read a book out loud to a child,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'laptop', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dishrack', 'door', 'light', 'shoe_rack']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to home office, walk to desk, sit in chair, turn on laptop, grab mouse, open a web page to look at it, move mouse, keep looking at laptop screen.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <mouse>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'tea_bag', 'bookshelf', 'mat', 'deck_of_cards', 'clothes_socks', 'closetdrawer', 'couch', 'mousepad', 'wall', 'nightstand', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'door', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Coming home turning on lights.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'faucet', 'mousepad', 'clothes_socks', 'dustpan', 'wall', 'coffe_maker', 'oven_mitts', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'sink', 'home_office', 'light', 'clothes_skirt', 'board_game', 'coffee']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the phone. I pick up the phone. I hold it to my ear.", "actions": ["[WALK] <dining_room>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'mop', 'bathroom', 'dough', 'milk', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'cleaning_solution', 'keyboard', 'mat', 'faucet', 'window', 'food_noodles', 'closetdrawer', 'couch', 'clothes_pants', 'nightstand', 'tape', 'food_food', 'check', 'food_rice', 'wooden_spoon', 'rag', 'mouse', 'controller', 'door', 'home_office']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'blow_dryer', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'dishwasher']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Turn on PC and log onto email account. Type email and send it. Switch off computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'shoes', 'door', 'table', 'light', 'creditcard']\nTask:Watch  TV\nresponse:\n{\n    \"thoughts\": Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'food_bread', 'dry_pasta', 'tablelamp', 'filing_cabinet', 'measuring_cup', 'microwave', 'toilet', 'chair', 'walllamp', 'floor', 'bookshelf', 'toaster', 'mat', 'faucet', 'shower', 'closetdrawer', 'bowl', 'dining_room', 'wall', 'newspaper', 'coffe_maker', 'oven_mitts', 'diary', 'kitchen_counter', 'cupboard', 'clothes_dress', 'sink', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen. Take the dishes from the sink. Apply dish washer, scrub and then wash with water.", "actions": ["[WALK] <sink>", "[FIND] <bowl>", "[GRAB] <bowl>", "[FIND] <dishwasher>", "[SCRUB] <dishwasher>", "[RINSE] <bowl>", "[OPEN] <dishwasher>", "[WASH] <bowl>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Draft home\nresponse:\n{\n    \"thoughts\": Using mathematics and CAD software design a home,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'cup', 'photoframe', 'television', 'tray', 'after_shave', 'doorjamb', 'coffee_pot', 'dresser', 'door', 'scrabble']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "watching TV from sofa", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[DROP] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'bookshelf', 'mat', 'shower', 'faucet', 'wall', 'powersocket', 'oven', 'light_bulb', 'novel', 'check', 'television', 'food_donut', 'curtain', 'doorjamb', 'kitchen_counter', 'spoon', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Switch on light in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[SWITCHON] <light>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'button', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'novel', 'television', 'doorjamb', 'dresser']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "first i will go to living room and then take the remote and swtitch on television after that il change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[TOUCH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'dough', 'orchid', 'cpuscreen', 'hanger', 'floor', 'keyboard', 'remote_control', 'food_cake', 'envelope', 'computer', 'button', 'food_bacon', 'ceilinglamp', 'stereo', 'homework', 'oven', 'tape', 'photoframe', 'coffee_table', 'keys', 'check', 'novel', 'light_bulb', 'wooden_spoon', 'curtain', 'cd', 'home_office']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_salt', 'bathroom', 'wallshelf', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'mat', 'oil', 'food_orange', 'bowl', 'wall', 'oven', 'powersocket', 'basket_for_clothes', 'check', 'after_shave', 'curtain', 'doorjamb', 'coffee_pot', 'dresser', 'pajamas', 'laundry_detergent', 'food_steak', 'sink', 'door', 'dishwasher', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'juice', 'chair', 'walllamp', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'powersocket', 'oven', 'television', 'coin', 'food_donut', 'doorjamb', 'mouse', 'dresser', 'maindoor', 'bathtub', 'light', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Turn on PC and log onto email account. Type email and send it. Switch off computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'photoframe', 'novel', 'television', 'mouse', 'kitchen_counter', 'door', 'sink', 'clothes_skirt', 'clothes_shirt']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to home office, find the book. Sit in the chair and read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'spectacles', 'food_ice_cream', 'crayon', 'microwave', 'toilet', 'band-aids', 'bathroom', 'needle', 'bathroom_counter', 'man', 'alcohol', 'mat', 'plate', 'stereo', 'faucet', 'mousepad', 'couch', 'pillow', 'dining_room', 'light_bulb', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'home_office', 'table', 'board_game']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": When someone knocks I n the door walk through the hallway and greet them.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['tea', 'walllamp', 'ground_coffee', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'bathroom_cabinet', 'table', 'scrabble', 'coffee', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Relax on the sofa in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'crayon', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'door', 'clothes_scarf', 'light']\nTask:Re arrange office\nresponse:\n{\n    \"thoughts\": Walk in.Turn on light.Move furniture. Change the room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[PUSH] <desk>', '[WALK] <chair>', '[PUSH] <chair>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'microwave', 'chair', 'wallshelf', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'mail', 'computer', 'toaster', 'mat', 'stovefan', 'faucet', 'mousepad', 'dining_room', 'wall', 'nightstand', 'coffe_maker', 'water', 'food_rice', 'drinking_glass', 'curtain', 'diary', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and approach the sink. I take a drinking glass and set it under the faucet. I turn the faucet until the water fills the glass, then turn it back. Finally I lift the glass out of the sink and drink the water.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[GRAB] <drinking_glass>", "[FIND] <water>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'shoes', 'door', 'table', 'light', 'creditcard']\nTask:Watch  TV\nresponse:\n{\n    \"thoughts\": Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'tea', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'laptop', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'bed', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to the home office. I put the laptop on the desk. I sit on the chair. I switch on my laptop and I type on the keyboard and browse the internet.", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <desk>", "[OPEN] <desk>", "[PUTBACK] <laptop> <desk>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'tablelamp', 'microwave', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'envelope', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'deck_of_cards', 'window', 'couch', 'wall', 'nightstand', 'novel', 'television', 'curtain', 'hairdryer', 'sink', 'home_office', 'table', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the television. Switch on the television. Find the remote control. Grab the remote control. Point at television. Find the button. Push button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'measuring_cup', 'dog', 'freezer', 'walllamp', 'orchid', 'bathroom_counter', 'keyboard', 'bookshelf', 'closetdrawer', 'pillow', 'dining_room', 'powersocket', 'oven', 'newspaper', 'clothes_gloves', 'bed', 'check', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'cd', 'cupboard', 'door', 'napkin', 'home_office', 'bathroom_cabinet', 'dishwasher', 'scrabble']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Read a book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'microwave', 'toilet', 'jelly', 'freezer', 'band-aids', 'drawing', 'knifeblock', 'clothes_underwear', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'powersocket', 'cup', 'basket_for_clothes', 'photoframe', 'food_food', 'water', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Get drink\nresponse:\n{\n    \"thoughts\": I go to the fridge and open the door. I look for the water pitcher, and take it out, making sure to close the fridge door behind me. Then I take a pint glass out of the upper cabinet over the sink and use the water pitcher to pour a full glass of water. Then I return the pitcher to the fridge and shut the door again, and take my glass of water out of the kitche.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[OPEN] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'toilet', 'bathroom_counter', 'floor', 'phone', 'desk', 'stovefan', 'toaster', 'plate', 'tvstand', 'picture', 'mat', 'mousepad', 'dining_room', 'food_kiwi', 'basket_for_clothes', 'cup', 'photoframe', 'bed', 'novel', 'keys', 'coin', 'food_rice', 'tray', 'pencil', 'kitchen_counter', 'door', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the cat. I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'knifeblock', 'knife', 'orchid', 'floor', 'ceiling', 'phone', 'keyboard', 'mail', 'stovefan', 'ceilinglamp', 'deck_of_cards', 'stereo', 'pillow', 'dining_room', 'wall', 'cup', 'bed', 'novel', 'coffe_maker', 'doorjamb', 'kitchen_counter', 'cupboard', 'controller', 'laundry_detergent', 'hairdryer', 'sink', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Making coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'bathroom', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'food_egg', 'bathroom_counter', 'shampoo', 'phone', 'mat', 'ceilinglamp', 'shower', 'electric_shaver', 'clothes_socks', 'couch', 'microphone', 'dining_room', 'food_kiwi', 'nightstand', 'wall', 'check', 'food_food', 'after_shave', 'mouse', 'kitchen_counter', 'cupboard', 'home_office', 'laser_pointer']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go home and it's dark so I can not see anything", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_bread', 'filing_cabinet', 'tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'mousepad', 'food_noodles', 'pillow', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'doorjamb', 'towel_rack', 'mouse', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to home office ,find computer ,switch on computer and work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'jelly', 'ground_coffee', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'phone', 'toaster', 'faucet', 'couch', 'pillow', 'wall', 'oven', 'light_bulb', 'novel', 'coffe_maker', 'razor', 'coffee_filter', 'food_rice', 'wine', 'kitchen_counter', 'cupboard', 'controller', 'sink', 'table', 'dishwasher', 'light', 'laser_pointer']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "You walk into the kitchen and turn to the coffee maker. You grab the coffee filter and put into the coffee machine, you also grab the ground coffee and put it into the coffee machine. You turn on the coffee machine.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'dog', 'cpuscreen', 'keyboard', 'ceilinglamp', 'plate', 'laptop', 'faucet', 'window', 'stereo', 'oil', 'food_noodles', 'mousepad', 'dining_room', 'nightstand', 'food_food', 'check', 'wine', 'curtain', 'after_shave', 'comb', 'dresser', 'napkin', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I reach out arm. I put hand on radio. I turn on radio. I pick a station.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[TOUCH] <stereo>', '[TURNTO] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'freezer', 'drawing', 'food_egg', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'desk', 'keyboard', 'stovefan', 'mat', 'shower', 'oil', 'couch', 'food_jam', 'pillow', 'wall', 'oven', 'powersocket', 'food_food', 'television', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into the kitchen. Open the fridge. Take out the chicken. Open the kitchen cabinet. Take out the frying pan. Put the chicken in the frying pan. Put the frying pan on the stove. Turn on the stove. Cook the chicken.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'detergent', 'cpuscreen', 'ceiling', 'computer', 'toaster', 'brush', 'laptop', 'faucet', 'mousepad', 'clothes_pants', 'dirt', 'powersocket', 'oven', 'cup', 'bed', 'food_food', 'television', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'chessboard', 'laundry_detergent', 'sink', 'dishwasher', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[PLUGIN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'envelope', 'mat', 'faucet', 'foundation', 'oil', 'couch', 'pillow', 'broom', 'wall', 'nightstand', 'food_food', 'television', 'razor', 'tray', 'curtain', 'centerpiece', 'cards', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'napkin']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'curtain', 'shoes', 'cd', 'vacuum_cleaner', 'dresser', 'cupboard', 'clothes_dress', 'home_office', 'table', 'shoe_rack']\nTask:Take off shoes\nresponse:\n{\n    \"thoughts\": Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'tea', 'microwave', 'chair', 'drawing', 'floor', 'character', 'toaster', 'laptop', 'plate', 'shower', 'tvstand', 'dish_soap', 'tooth_paste', 'dustpan', 'candle', 'bowl', 'wall', 'cup', 'coffe_maker', 'check', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'sink', 'instrument_guitar', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Using dishwasher to wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'mop', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'candle', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'check', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'door', 'bathroom_cabinet', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I go to the living room, switch off the light, sit in the sofa and do nothing.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHOFF] <light>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['food_salt', 'cat', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'oil', 'closetdrawer', 'food_orange', 'form', 'pillow', 'wall', 'nightstand', 'powersocket', 'tape', 'photoframe', 'doorjamb', 'console', 'kitchen_counter', 'door', 'light', 'dishwasher', 'clothes_skirt', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "approach the light switch, raise hand to the switch, flip the switch", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'pillow', 'wall', 'nightstand', 'photoframe', 'wine', 'doorjamb', 'mouse', 'dresser', 'napkin', 'bathroom_cabinet', 'light']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Wipe down computer desk and monitor,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <desk>', '[FIND] <computer>', '[WIPE] <computer>', '[WIPE] <desk>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'beer', 'chair', 'freezer', 'drawing', 'wallshelf', 'hanger', 'phone', 'dvd_player', 'purse', 'clothes_jacket', 'mat', 'stereo', 'oil', 'pillow', 'nightstand', 'cup', 'newspaper', 'keys', 'bed', 'check', 'video_game_controller', 'television', 'lighter', 'chessboard', 'dresser', 'napkin', 'spoon', 'table', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to Living Room. Find table. Walk to table. Find telephone. Lift telephone.", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[WALK] <table>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'orchid', 'knifeblock', 'wallshelf', 'character', 'keyboard', 'stovefan', 'mat', 'closetdrawer', 'dining_room', 'food_kiwi', 'box', 'cup', 'bed', 'coffe_maker', 'television', 'water', 'colander', 'tray', 'curtain', 'kitchen_counter', 'cd', 'cupboard', 'door', 'bag', 'soap', 'creditcard']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": put soap on hands, scrub it and rinse with water,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'shoe_rack']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the chair. I find the television remote control. I turn it on. I find the television. I look at the television. I watched television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'bed', 'coin', 'curtain', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'light']\nTask:Wipe down baseboards please\nresponse:\n{\n    \"thoughts\": Use a rag to get all the dust off the baseboards please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <wall>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'food_pizza', 'pot', 'microwave', 'toilet', 'nail_polish', 'orchid', 'stamp', 'desk', 'food_cake', 'toaster', 'oil', 'candle', 'broom', 'nightstand', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'novel', 'coffe_maker', 'mouse', 'kitchen_counter', 'cd', 'shoes', 'food_steak', 'clothes_scarf', 'home_office', 'clothes_skirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into living room, look at the phone when it ringing, grab the phone or pickup the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'clothes_socks', 'couch', 'pillow', 'wall', 'powersocket', 'box', 'cup', 'photoframe', 'television', 'food_rice', 'doorjamb', 'towel_rack', 'kitchen_counter', 'door', 'home_office', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'ground_coffee', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'folder', 'purse', 'computer', 'toaster', 'laptop', 'shower', 'picture', 'couch', 'wall', 'pasta', 'nightstand', 'box', 'photoframe', 'coffe_maker', 'television', 'coffee_filter', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'instrument_violin']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open the coffee maker and put a filer, grab ground coffee and put it in the coffee maker, then close the coffee maker and switch it on", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'juice', 'toilet', 'chair', 'nail_polish', 'needle', 'clothes_underwear', 'phone', 'keyboard', 'bookmark', 'stovefan', 'ceilinglamp', 'toaster', 'shower', 'foundation', 'clothes_hat', 'pillow', 'oven', 'powersocket', 'tape', 'bed', 'coffe_maker', 'television', 'curtain', 'towel_rack', 'mouse', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the radio. Turn on the radio. Listen to the radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'measuring_cup', 'mop', 'freezer', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'phone', 'bedroom', 'purse', 'mat', 'plate', 'stereo', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'mouse', 'kitchen_counter', 'cd', 'dresser', 'light', 'dishwasher', 'soap', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into my bedroom and see my stereo. I decide to play music. I walk over to the stereo and turn it on. I sit down on the bed to listen to the music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'console', 'dresser', 'cupboard', 'home_office', 'table', 'clothes_skirt', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'bathroom', 'detergent', 'food_carrot', 'cpuscreen', 'wallshelf', 'phone', 'bedroom', 'desk', 'remote_control', 'envelope', 'computer', 'mail', 'tvstand', 'shower', 'iron', 'mousepad', 'closetdrawer', 'oven', 'powersocket', 'newspaper', 'novel', 'food_food', 'curtain', 'dresser', 'food_steak', 'sink', 'table', 'shoe_rack']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[PUSH] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'tablelamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'purse', 'computer', 'clothes_jacket', 'toaster', 'stereo', 'mousepad', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'bed', 'novel', 'mouse', 'shoes', 'dresser', 'clothes_dress', 'light', 'table']\nTask:Get ready for day\nresponse:\n{\n    \"thoughts\": Grab shoes, purse etc and head out the door!,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTON] <clothes_dress>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>', '[FIND] <purse>', '[GRAB] <purse>', '[CLOSE] <dresser>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'hair', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to home office and work on computer.", "actions": ["[RUN] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TOUCH] <mouse>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'crayon', 'toilet', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'mail', 'food_bacon', 'clothes_jacket', 'table_cloth', 'faucet', 'mousepad', 'closetdrawer', 'food_jam', 'nightstand', 'powersocket', 'novel', 'food_food', 'centerpiece', 'diary', 'console', 'cupboard', 'bathroom_cabinet', 'table', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Pick up the ringing phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[POINTAT] <table>', '[FIND] <phone>', '[GRAB] <phone>', '[PUTOBJBACK] <phone>']\n}\n", "input": "\nScene_Objects:['blow_dryer', 'bathroom', 'food_carrot', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stamp', 'bedroom', 'food_cake', 'computer', 'shower', 'picture', 'oven', 'basket_for_clothes', 'photoframe', 'food_food', 'check', 'towel_rack', 'mouse', 'pencil', 'shoes', 'comb', 'hairbrush', 'cupboard', 'laundry_detergent', 'sink', 'napkin', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Pick up my dirty clothes, load the washer, add detergent, turn on washer", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[RUN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'shower', 'window', 'homework', 'mousepad', 'pillow', 'powersocket', 'photoframe', 'coffe_maker', 'doorjamb', 'kettle', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'clothes_dress', 'door', 'bag', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to bedroom and grab dirty pants and take to the washing machine in entrance hall. Put clothes in washing machine and then put in laundry detergent.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_pants>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'measuring_cup', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'coffe_maker', 'television', 'tray', 'doorjamb', 'dresser', 'coffee']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk to living room ,find television ,find remote control ,switch on remote control and change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[SWITCHON] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'chair', 'freezer', 'food_butter', 'drawing', 'bills', 'desk', 'tea_bag', 'keyboard', 'purse', 'food_cake', 'envelope', 'plate', 'dish_soap', 'mousepad', 'pillow', 'wall_clock', 'powersocket', 'shelf', 'standingmirror', 'check', 'coffe_maker', 'doorjamb', 'towel_rack', 'shaving_cream', 'kitchen_counter', 'dresser', 'sink']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Pick up dish.  Put in soapy water.  Wash dish.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dish_soap>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <dish_soap> <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WIPE] <plate>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'toilet', 'chair', 'ground_coffee', 'bathroom', 'dough', 'wallshelf', 'phone', 'character', 'keyboard', 'computer', 'ceilinglamp', 'deck_of_cards', 'stereo', 'mousepad', 'couch', 'powersocket', 'cutting_board', 'bed', 'food_food', 'food_rice', 'tray', 'wooden_spoon', 'pencil', 'headset', 'cupboard', 'laundry_detergent', 'sink', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find the cat, go near the cat and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_salt', 'beer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'bookshelf', 'mail', 'computer', 'button', 'toaster', 'plate', 'shower', 'dish_soap', 'mousepad', 'pillow', 'wall', 'oven', 'photoframe', 'clothes_gloves', 'coffe_maker', 'tray', 'food_onion', 'doorjamb', 'dresser', 'hairdryer', 'sink', 'dishwasher', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put dishes in the dishwasher and start the dishwasher.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[WALK] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'tablelamp', 'blender', 'tea', 'toothbrush', 'cpuscreen', 'hanger', 'bedroom', 'computer', 'laptop', 'plate', 'faucet', 'candle', 'food_jam', 'wall_clock', 'bed', 'food_food', 'check', 'television', 'food_donut', 'after_shave', 'lighter', 'cards', 'mouse', 'kitchen_counter', 'dresser', 'pajamas', 'laundry_detergent', 'home_office']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go to the living room where the cat sleeps. Find cat. Proceed to petting cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'tea_bag', 'ceilinglamp', 'tvstand', 'shower', 'window', 'homework', 'couch', 'dustpan', 'food_orange', 'dining_room', 'wall', 'nightstand', 'photoframe', 'clothes_gloves', 'coffe_maker', 'check', 'television', 'mouse', 'pencil', 'table', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, turn to the sofa. Sit on the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'dvd_player', 'keyboard', 'stovefan', 'mat', 'laptop', 'stereo', 'oil', 'mousepad', 'couch', 'pillow', 'wall', 'food_food', 'television', 'drinking_glass', 'water', 'curtain', 'diary', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'light', 'instrument_violin']\nTask:Drink\nresponse:\n", "output": {"thoughts": "go to the kitchen", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <water>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'chair', 'bathroom', 'drawing', 'food_egg', 'floor', 'desk', 'bookshelf', 'computer', 'mat', 'stereo', 'shower', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'cup', 'doorjamb', 'mouse', 'kitchen_counter', 'pajamas', 'controller', 'door', 'clothes_scarf', 'light', 'dishwasher', 'board_game']\nTask:Pull up carpet\nresponse:\n{\n    \"thoughts\": Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mat>', '[FIND] <mat>', '[TOUCH] <mat>', '[PULL] <mat>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'food_carrot', 'drawing', 'hanger', 'phone', 'ceiling', 'desk', 'computer', 'toaster', 'mat', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'towel', 'cup', 'food_cheese', 'keys', 'tray', 'after_shave', 'towel_rack', 'mouse', 'cupboard', 'sink', 'coffee']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room and I pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'ground_coffee', 'nail_polish', 'orchid', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'food_cake', 'bookshelf', 'mail', 'toaster', 'plate', 'homework', 'pillow', 'dining_room', 'wall', 'bed', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[GRAB] <ground_coffee>', '[PUTOBJBACK] <ground_coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'freezer', 'bathroom', 'cat', 'milk', 'ceiling', 'desk', 'bedroom', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'oil', 'closetdrawer', 'clothes_pants', 'food_jam', 'broom', 'wall', 'basket_for_clothes', 'food_food', 'razor', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'soap', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'food_cake', 'stovefan', 'mat', 'toaster', 'shower', 'food_noodles', 'oil', 'nightstand', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'laptop', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I read a book in my office", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'food_ice_cream', 'microwave', 'toilet', 'freezer', 'ceilingfan', 'knife', 'character', 'desk', 'bookshelf', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'couch', 'candle', 'clothes_pants', 'food_orange', 'wall', 'nightstand', 'bench', 'tape', 'check', 'curtain', 'doorjamb', 'dishrack', 'clothes_dress', 'food_steak', 'scrabble', 'coffee']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Pick up the book, open to current page, read the words.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'tea_bag', 'computer', 'mat', 'tvstand', 'mousepad', 'couch', 'candle', 'pillow', 'bowl', 'wall', 'oven', 'toy', 'cup', 'photoframe', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'dresser', 'laundry_detergent', 'sink', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room, find the sofa and sit in the sofa and find pillow. lie down in sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <pillow>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'freezer', 'trashcan', 'milk', 'knifeblock', 'orchid', 'clothes_underwear', 'bookshelf', 'purse', 'thread', 'computer', 'ceilinglamp', 'tvstand', 'iron', 'couch', 'microphone', 'wall', 'cutting_board', 'basket_for_clothes', 'keys', 'toothbrush_holder', 'oven_mitts', 'curtain', 'doorjamb', 'diary', 'chessboard', 'home_office', 'bathroom_cabinet', 'clothes_skirt', 'light']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'band-aids', 'nail_polish', 'toothbrush', 'drawing', 'bathroom_counter', 'phone', 'food_cake', 'ceilinglamp', 'shoe-shine_kit', 'mousepad', 'closetdrawer', 'dining_room', 'wall_clock', 'food_cheese', 'newspaper', 'novel', 'television', 'coffee_filter', 'tray', 'after_shave', 'rag', 'centerpiece', 'curtain', 'console', 'kitchen_counter', 'clothes_dress', 'bathroom_cabinet', 'dishwasher']\nTask:Read book\nresponse:\n", "output": {"thoughts": "walk to living room ,find a book ,open the book ,read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'dining_room', 'wall', 'powersocket', 'oven', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Push power button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'razor', 'food_onion', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I walk in the living room, find the computer and sit down on a chair. I turn on the computer. I grab the computer mouse.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_salt', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'desk', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'clothes_pants', 'dining_room', 'wall', 'toy', 'basket_for_clothes', 'bed', 'television', 'coin', 'coffee_filter', 'food_rice', 'curtain', 'cupboard', 'controller', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch TV in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'dog', 'cat', 'drawing', 'wallshelf', 'ceiling', 'folder', 'character', 'cleaning_solution', 'envelope', 'computer', 'brush', 'toaster', 'iron', 'mousepad', 'closetdrawer', 'couch', 'bed', 'photoframe', 'novel', 'coffe_maker', 'centerpiece', 'hair', 'cupboard', 'sink', 'home_office', 'light', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[SQUEEZE] <cleaning_solution>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'nail_polish', 'knifeblock', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'desk', 'cleaning_solution', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'shower', 'faucet', 'ceilinglamp', 'pillow', 'bench', 'cup', 'mouse', 'controller', 'laundry_detergent', 'door', 'napkin', 'light', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'detergent', 'cat', 'drawing', 'bills', 'wallshelf', 'cpuscreen', 'phone', 'character', 'shower', 'food_orange', 'form', 'pillow', 'nightstand', 'photoframe', 'food_food', 'tray', 'lighter', 'shaving_cream', 'mouse', 'towel_rack', 'dishrack', 'laundry_detergent', 'spoon', 'bathroom_cabinet', 'light', 'home_office', 'soap', 'instrument_violin', 'shoe_rack']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go into the living room and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'chair', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'bedroom', 'button', 'plate', 'dish_soap', 'wall', 'nightstand', 'food_cheese', 'photoframe', 'newspaper', 'coffe_maker', 'check', 'tray', 'kettle', 'pencil', 'kitchen_counter', 'comb', 'cupboard', 'door', 'table', 'dishwasher', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'food_salt', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'envelope', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'centerpiece', 'dresser', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room. Sit on the sofa. Grab the remote control and plug in TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[PLUGIN] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'dry_pasta', 'filing_cabinet', 'scissors', 'kitchen_cabinet', 'envelope', 'faucet', 'dish_soap', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall_clock', 'bench', 'toy', 'cutting_board', 'razor', 'food_onion', 'doorjamb', 'towel_rack', 'kitchen_counter', 'pajamas', 'clothes_dress', 'bathtub', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'vase', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTIN] <dish_soap> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'stovefan', 'mat', 'tvstand', 'iron', 'ceilinglamp', 'oil', 'couch', 'dustpan', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'dresser', 'pajamas', 'hairdryer', 'light', 'soap', 'dirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sponge', 'food_bread', 'blender', 'chair', 'walllamp', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'tvstand', 'window', 'mousepad', 'clothes_socks', 'wall', 'oven', 'cup', 'television', 'mouse', 'cupboard', 'door', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": go to the office, turn computer on, open an browser or any application to open any website.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <computer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'tablelamp', 'food_salt', 'microwave', 'chair', 'bathroom', 'detergent', 'tea_bag', 'keyboard', 'ceilinglamp', 'plate', 'dining_room', 'food_kiwi', 'nightstand', 'powersocket', 'basket_for_clothes', 'check', 'mouthwash', 'doorjamb', 'towel_rack', 'hairbrush', 'pajamas', 'door', 'home_office', 'table', 'soap', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "put clothes and soap in the washing machine, turn it on", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'bookshelf', 'ceilinglamp', 'tvstand', 'shower', 'mat', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'oven', 'coffe_maker', 'tray', 'curtain', 'coffee_pot', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'hairdryer']\nTask:Answer door\nresponse:\n{\n    \"thoughts\": Walk up to the door.  Open the door and greet whomever is out side,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['sponge', 'scissors', 'chair', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'mail', 'bookmark', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'dish_soap', 'bowl', 'wall', 'basket_for_clothes', 'coffe_maker', 'food_food', 'curtain', 'doorjamb', 'kettle', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'spoon', 'table']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I enter the kitchen, walk to the sink, turn on the faucet, grab the sponge and get it wet, then apply dish soap to the sponge. I take a dish, scrub it with the sponge, rinse it, and place it on the dish rack to dry, and repeat until all the dishes are clean. I rinse the sponge, put it down, and turn off the faucet.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <sponge>", "[FIND] <dish_soap>", "[GRAB] <sponge>", "[RINSE] <sponge>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[SQUEEZE] <sponge>", "[FIND] <bowl>", "[GRAB] <bowl>", "[SCRUB] <bowl>", "[RINSE] <bowl>", "[FIND] <dishrack>", "[PUTBACK] <bowl> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <spoon>", "[GRAB] <spoon>", "[SCRUB] <spoon>", "[RINSE] <spoon>", "[PUTBACK] <spoon> <dishrack>", "[RINSE] <sponge>", "[SQUEEZE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'tablelamp', 'food_ice_cream', 'freezer', 'bathroom', 'orchid', 'bills', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'stereo', 'closetdrawer', 'nightstand', 'oven', 'photoframe', 'clothes_gloves', 'toilet_paper', 'video_game_controller', 'television', 'doorjamb', 'mouse', 'cupboard', 'controller', 'door', 'spoon', 'bathroom_cabinet', 'light', 'dishwasher', 'instrument_violin']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'blow_dryer', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'tvstand', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'photoframe', 'novel', 'food_food', 'television', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'coffee_cup', 'table', 'board_game']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'drawing', 'floor', 'purse', 'computer', 'plate', 'balanceball', 'shower', 'stereo', 'dustpan', 'clothes_hat', 'microphone', 'bench', 'basket_for_clothes', 'cup', 'bed', 'check', 'coffe_maker', 'food_onion', 'after_shave', 'cards', 'pencil', 'chessboard', 'maindoor', 'napkin', 'table', 'board_game', 'clothes_shirt', 'coffee']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'check', 'television', 'ice', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk to living room.grab remote control. look at television. switch on television. lift remote control. push remote control. watch television. switch off television. put back remote control.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>", "[PUSH] <remote_control>", "[WATCH] <television>", "[SWITCHOFF] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'juice', 'microwave', 'freezer', 'blow_dryer', 'ground_coffee', 'cat', 'floor', 'legs_both', 'face', 'toaster', 'laptop', 'homework', 'closetdrawer', 'towel', 'feet_both', 'arms_both', 'basket_for_clothes', 'newspaper', 'photoframe', 'coffe_maker', 'mouthwash', 'centerpiece', 'doorjamb', 'headset', 'dresser', 'cupboard', 'napkin', 'soap', 'clothes_shirt']\nTask:Try yourself off\nresponse:\n{\n    \"thoughts\": After your shower, go into your bedroom and use a towel to dry water off of yourself before dressing.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <dresser>', '[FIND] <face>', '[WIPE] <face>', '[FIND] <arms_both>', '[WIPE] <arms_both>', '[FIND] <legs_both>', '[WIPE] <legs_both>', '[FIND] <feet_both>', '[WIPE] <feet_both>', '[FIND] <basket_for_clothes>', '[PUTBACK] <towel> <basket_for_clothes>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'dough', 'wallshelf', 'hanger', 'floor', 'phone', 'folder', 'desk', 'tea_bag', 'mousepad', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'bed', 'television', 'toothbrush_holder', 'food_onion', 'curtain', 'lighter', 'diary', 'cd', 'dresser', 'controller', 'sink', 'table', 'light', 'dishwasher']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to home office 1. Walk over to phone 1 on table 1.  Grab phone 1.  lift phone 1.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'toothbrush', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'thread', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'plate', 'dining_room', 'wall', 'wall_clock', 'oven', 'toothbrush_holder', 'tray', 'doorjamb', 'towel_rack', 'chessboard', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": getting ready for bed,\n    \"actions\": ['[WALK] <bathroom_counter>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <toothbrush> <bathroom_counter>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>', '[GRAB] <toothbrush>', '[SCRUB] <toothbrush>', '[RINSE] <toothbrush>', '[FIND] <toothbrush_holder>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'envelope', 'computer', 'button', 'food_bacon', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'shoe-shine_kit', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'photoframe', 'television', 'doorjamb', 'dresser', 'dirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Pickup remote control for the living room TV and push the change channel button", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'knifeblock', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'desk', 'food_cake', 'stovefan', 'toaster', 'deck_of_cards', 'faucet', 'candle', 'pillow', 'dining_room', 'oven', 'photoframe', 'check', 'food_food', 'television', 'bathroom_cabinet', 'table', 'dishwasher', 'soap', 'board_game', 'clothes_shirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find cat. Walk to cat. Pull cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[PULL] <cat>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'microwave', 'dog', 'detergent', 'knife', 'bathroom_counter', 'floor', 'toaster', 'tvstand', 'shower', 'faucet', 'dish_soap', 'mousepad', 'closetdrawer', 'couch', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'oven_mitts', 'wine', 'tray', 'centerpiece', 'kitchen_counter', 'dresser', 'cupboard', 'clothes_dress', 'door', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk to kitchen, find the dishwasher, apply dish soap into vessels or dishwasher, wash the dishes with a dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[WASH] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'ground_coffee', 'drawing', 'orchid', 'cpuscreen', 'phone', 'bedroom', 'stamp', 'desk', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'nightstand', 'wall', 'oven', 'wall_clock', 'photoframe', 'coffe_maker', 'mouse', 'dresser', 'cupboard', 'bathtub', 'sink', 'home_office', 'table', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": You walk to the living room and find the cat. You look at the cat and touch it to pet it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'newspaper', 'television', 'food_donut', 'food_rice', 'doorjamb', 'mouse', 'cd', 'dresser']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "turn on the tv", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'chair', 'bathroom', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'phone', 'character', 'computer', 'toaster', 'mat', 'tvstand', 'couch', 'clothes_hat', 'pillow', 'wall', 'wall_clock', 'bench', 'powersocket', 'cup', 'bed', 'food_food', 'television', 'mouse', 'pencil', 'bathroom_cabinet', 'table']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": walk to entrance hall. find a table. pull table. walk to living rook. push table.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[PULL] <table>', '[WALK] <home_office>', '[WALK] <table>', '[PUSH] <table>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'form', 'pillow', 'wall', 'pasta', 'powersocket', 'photoframe', 'television', 'food_donut', 'curtain', 'doorjamb', 'chessboard', 'dresser', 'laundry_detergent', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'bedroom', 'mail', 'stovefan', 'faucet', 'stereo', 'couch', 'wall', 'oven', 'cutting_board', 'bed', 'newspaper', 'coffe_maker', 'food_food', 'tray', 'lighter', 'doorjamb', 'kitchen_counter', 'home_office', 'bathroom_cabinet', 'vase', 'soap']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Open fridge. Put groceries in fridge.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'jelly', 'freezer', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'food_cake', 'mop_bucket', 'stovefan', 'mat', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'food_food', 'television', 'wine', 'tray', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'light', 'clothes_skirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen, open fridge, get groceries, put groceries in fridge, close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'walllamp', 'nail_polish', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bookshelf', 'computer', 'faucet', 'oil', 'dining_room', 'wall', 'nightstand', 'oven', 'powersocket', 'toy', 'cup', 'coffee_filter', 'shaving_cream', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'home_office', 'table', 'dishwasher', 'bathroom_cabinet']\nTask:Get glass of water\nresponse:\n{\n    \"thoughts\": Get a glass from a cabinet. Hold it under the sink and turn the cold water on. Fill the glass. Turn off the water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <kitchen_cabinet>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'mop_bucket', 'toaster', 'mat', 'tvstand', 'stereo', 'shower', 'deck_of_cards', 'homework', 'electric_shaver', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'newspaper', 'television', 'doorjamb', 'cd', 'bathroom_cabinet', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room. Walk to radio. Switch on the radio. Walk to sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'beer', 'chair', 'bills', 'knife', 'hanger', 'bathroom_counter', 'floor', 'computer', 'stovefan', 'mat', 'tvstand', 'window', 'couch', 'pillow', 'towel', 'dining_room', 'food_kiwi', 'wall', 'toy', 'photoframe', 'novel', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'bag', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on couch. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <couch>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'spectacles', 'tablelamp', 'chair', 'walllamp', 'bills', 'drawing', 'wallshelf', 'cpuscreen', 'orchid', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I walk into my office and go to the computer desk. I sit at my desk and turn the computer on. Then I write an email.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'trashcan', 'knifeblock', 'floor', 'character', 'computer', 'table_cloth', 'toaster', 'plate', 'faucet', 'dish_soap', 'wall', 'oven', 'bench', 'cup', 'coffe_maker', 'food_food', 'television', 'wine', 'tray', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['fork', 'water_glass', 'pot', 'tablelamp', 'microwave', 'drawing', 'knife', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'toaster', 'stovefan', 'plate', 'shower', 'dish_soap', 'mousepad', 'wall', 'oven', 'photoframe', 'coffe_maker', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'spoon', 'coffee_cup', 'dishwasher', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and put plates, cutlery and glasses into the dishwasher, put the soap, close the machine and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'mat', 'ceilinglamp', 'shower', 'plate', 'wall', 'nightstand', 'powersocket', 'food_cheese', 'bed', 'coffe_maker', 'tray', 'curtain', 'towel_rack', 'dresser', 'controller', 'clothes_scarf', 'home_office', 'bathroom_cabinet', 'board_game', 'scrabble', 'teeth']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": walk to bathroom , find shower , take a shower,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'tablelamp', 'food_carrot', 'toothbrush', 'drawing', 'wallshelf', 'kitchen_cabinet', 'floor', 'keyboard', 'alcohol', 'stovefan', 'tvstand', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'nightstand', 'wall', 'powersocket', 'box', 'keys', 'coffe_maker', 'water', 'tray', 'kitchen_counter', 'clothes_dress', 'sink', 'bathroom_cabinet', 'board_game']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to kitchen. Open the the cabinet. Locate the water glass. Pour water into glass. Drag the bottle close to mouth. Drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[CLOSE] <kitchen_cabinet>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'sponge', 'crayon', 'microwave', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'floor', 'bookshelf', 'button', 'stovefan', 'mat', 'stereo', 'mousepad', 'clothes_hat', 'bowl', 'nightstand', 'wall', 'bed', 'photoframe', 'light_bulb', 'food_food', 'television', 'kitchen_counter', 'dresser', 'door', 'sink', 'clothes_skirt', 'scrabble']\nTask:Breakfast\nresponse:\n{\n    \"thoughts\": I have breakfast in bed, watching tv.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[GRAB] <food_food>', '[WALK] <bedroom>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <bed>', '[LIE] <bed>', '[TURNTO] <television>', '[WATCH] <television>', '[EAT] <food_food>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'filing_cabinet', 'microwave', 'chair', 'jelly', 'bathroom', 'detergent', 'knife', 'floor', 'toaster', 'mat', 'plate', 'shower', 'faucet', 'dish_soap', 'bowl', 'wall', 'cup', 'coffe_maker', 'food_food', 'after_shave', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'spoon', 'home_office', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the dishwasher. I open it. I find the sink and grab the dirty plates, cups, bowls, and utensils. I rinse off the dirty dishes under the faucet. I then place the dirty dishes into the dishwasher. I add soap to the dishwasher and close the lid. I turn it on and wait.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <sink>", "[FIND] <bowl>", "[GRAB] <bowl>", "[RINSE] <bowl>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[RINSE] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[RINSE] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[RINSE] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[RINSE] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[TURNTO] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'microwave', 'chair', 'walllamp', 'trashcan', 'food_butter', 'orchid', 'floor', 'dvd_player', 'desk', 'toaster', 'tvstand', 'oil', 'wall', 'bench', 'basket_for_clothes', 'photoframe', 'coffe_maker', 'razor', 'wooden_spoon', 'curtain', 'coffee_pot', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'spoon', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <dishrack>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <dishrack> <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'cat', 'dough', 'bathroom_counter', 'phone', 'bookshelf', 'ceilinglamp', 'tvstand', 'stereo', 'candle', 'bowl', 'dining_room', 'wall', 'newspaper', 'coffe_maker', 'food_food', 'coffee_filter', 'tray', 'curtain', 'diary', 'towel_rack', 'pencil', 'dresser', 'food_steak', 'sink', 'napkin', 'door', 'dishwasher']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I have my desktop computer set up on my desk and complete the paperwork I brought home with me from work.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <check> <desk>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'filing_cabinet', 'blender', 'toilet', 'drawing', 'needle', 'food_egg', 'clothes_underwear', 'bathroom_counter', 'phone', 'keyboard', 'computer', 'table_cloth', 'closetdrawer', 'clothes_hat', 'nightstand', 'oven', 'basket_for_clothes', 'photoframe', 'check', 'television', 'coin', 'food_donut', 'diary', 'towel_rack', 'headset', 'comb', 'hairdryer', 'napkin']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drink water", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'needle', 'floor', 'ceiling', 'phone', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'stereo', 'iron', 'stovefan', 'couch', 'pillow', 'wall', 'television', 'food_rice', 'curtain', 'doorjamb', 'bag', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'shampoo', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "I go to my office. I switch the computer on and sit on the office chair. I look at the computer screen to see what needs to be done. I use my mouse to navigate to the relevant directory, then release it. I type code with my keyboard. When I'm done, I look at the computer again to evaluate what I've done. Then I switch off the computer using the mouse and put back the mouse.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[GRAB] <mouse>", "[SWITCHOFF] <computer>", "[PUTOBJBACK] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'orchid', 'hanger', 'floor', 'phone', 'ceiling', 'character', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'window', 'closetdrawer', 'powersocket', 'oven', 'cup', 'curtain', 'doorjamb', 'mouse', 'comb', 'cupboard', 'clothes_dress', 'door', 'bathtub', 'home_office', 'table', 'dishwasher', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'crayon', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'ceilinglamp', 'clothes_socks', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'food_food', 'video_game_controller', 'doorjamb', 'diary', 'chessboard', 'dresser', 'door', 'light', 'table', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "flip the switch and turn on the lights", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'bathroom', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'food_bacon', 'button', 'toaster', 'tvstand', 'shower', 'plate', 'food_noodles', 'clothes_socks', 'dustpan', 'pillow', 'nightstand', 'pasta', 'wall_clock', 'bed', 'novel', 'food_donut', 'towel_rack', 'kitchen_counter', 'comb', 'cupboard', 'door', 'table', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": go to radio. run on radio.,\n    \"actions\": ['[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'milk', 'bathroom_counter', 'floor', 'bookshelf', 'table_cloth', 'toaster', 'tvstand', 'shower', 'foundation', 'pillow', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'cup', 'coffe_maker', 'food_onion', 'kitchen_counter', 'dishrack', 'cupboard', 'clothes_dress', 'sink', 'bag', 'table', 'dishwasher', 'dirt', 'shoe_rack']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dishrack>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <dishrack> <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'toilet', 'chair', 'ground_coffee', 'bathroom', 'dough', 'wallshelf', 'phone', 'character', 'keyboard', 'computer', 'ceilinglamp', 'deck_of_cards', 'stereo', 'mousepad', 'couch', 'powersocket', 'cutting_board', 'bed', 'food_food', 'food_rice', 'tray', 'wooden_spoon', 'pencil', 'headset', 'cupboard', 'laundry_detergent', 'sink', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find the cat, go near the cat and pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to over to desk, sit down on chair and face the computer, turn the computer on, use arms to place hands on the keyboard, type work into the computer and use eyes to review.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'ceilingfan', 'orchid', 'bathroom_counter', 'bookshelf', 'purse', 'stovefan', 'mat', 'iron', 'electric_shaver', 'food_noodles', 'mousepad', 'couch', 'clothes_pants', 'pillow', 'powersocket', 'keys', 'coffe_maker', 'oven_mitts', 'wine', 'shaving_cream', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'home_office', 'light', 'tooth_paste']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'walllamp', 'food_carrot', 'knife', 'orchid', 'cpuscreen', 'phone', 'dvd_player', 'desk', 'mop_bucket', 'toaster', 'laptop', 'tvstand', 'closetdrawer', 'food_jam', 'pillow', 'dining_room', 'food_food', 'television', 'oven_mitts', 'water', 'after_shave', 'pencil', 'hairbrush', 'clothes_dress', 'door', 'table', 'dishwasher', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'freezer', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'computer', 'stovefan', 'tvstand', 'faucet', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'curtain', 'rag', 'doorjamb', 'towel_rack', 'sink', 'bathroom_cabinet', 'table', 'light']\nTask:Wipe down sink\nresponse:\n{\n    \"thoughts\": Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[WIPE] <sink>', '[PUTOBJBACK] <rag>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['chair', 'bathroom', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'food_bacon', 'button', 'toaster', 'tvstand', 'shower', 'plate', 'food_noodles', 'clothes_socks', 'dustpan', 'pillow', 'nightstand', 'pasta', 'wall_clock', 'bed', 'novel', 'food_donut', 'towel_rack', 'kitchen_counter', 'comb', 'cupboard', 'door', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "go to radio. run on radio.", "actions": ["[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blow_dryer', 'trashcan', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookmark', 'toaster', 'stovefan', 'tvstand', 'food_jam', 'pillow', 'dining_room', 'bowl', 'wall', 'powersocket', 'clothes_gloves', 'novel', 'coffe_maker', 'coffee_filter', 'tray', 'after_shave', 'curtain', 'cd', 'chessboard', 'cupboard', 'sink', 'dishwasher', 'scrabble', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[WALK] <coffee>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'purse', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'electric_shaver', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office. Sit on chair. Switch on computer. type in keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'toaster', 'stovefan', 'plate', 'faucet', 'iron', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'food_food', 'wooden_spoon', 'tray', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'headset', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <chair>", "[LOOKAT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'toilet', 'chair', 'freezer', 'detergent', 'food_butter', 'wallshelf', 'hanger', 'phone', 'character', 'desk', 'remote_control', 'button', 'piano_bench', 'toaster', 'mat', 'tvstand', 'picture', 'oil', 'couch', 'clothes_pants', 'pillow', 'towel', 'wall', 'novel', 'razor', 'diary', 'conditioner', 'table', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'keys', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "go to the office, turn computer on, open an browser or any application to open any website.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'detergent', 'knife', 'bills', 'bathroom_counter', 'phone', 'folder', 'bedroom', 'character', 'computer', 'toaster', 'stovefan', 'mat', 'window', 'mousepad', 'pillow', 'oven', 'check', 'coffe_maker', 'razor', 'after_shave', 'mouse', 'kitchen_counter', 'shoes', 'dresser', 'pajamas', 'sink', 'bag']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": I get dressed in my bedroom in the mornings.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n", "output": {"thoughts": "Find home office. Walk to home office. Find computer. Find chair at desk. Sit in chair. Switch on computer. Work on computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'juice', 'microwave', 'chair', 'freezer', 'detergent', 'drawing', 'orchid', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'shower', 'mat', 'ceilinglamp', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'newspaper', 'coffe_maker', 'television', 'ice', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'cupboard', 'table', 'dishwasher', 'creditcard']\nTask:Grab some juice\nresponse:\n{\n    \"thoughts\": I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'crayon', 'toilet', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'mail', 'food_bacon', 'clothes_jacket', 'table_cloth', 'faucet', 'mousepad', 'closetdrawer', 'food_jam', 'nightstand', 'powersocket', 'novel', 'food_food', 'centerpiece', 'diary', 'console', 'cupboard', 'bathroom_cabinet', 'table', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Pick up the ringing phone", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[TURNTO] <table>", "[POINTAT] <table>", "[FIND] <phone>", "[GRAB] <phone>", "[PUTOBJBACK] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'microwave', 'chair', 'detergent', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'purse', 'computer', 'toaster', 'mat', 'tvstand', 'clothes_socks', 'food_peanut_butter', 'microphone', 'food_jam', 'pillow', 'wall', 'powersocket', 'after_shave', 'doorjamb', 'cd', 'pajamas', 'laundry_detergent', 'door', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Coming home turning on lights.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'envelope', 'computer', 'table_cloth', 'mat', 'tvstand', 'iron', 'wall', 'powersocket', 'oven', 'bed', 'television', 'tray', 'after_shave', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'door', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find light switch.  Switch on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'chair', 'detergent', 'toothbrush', 'floor', 'ceiling', 'food_cake', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'ceilinglamp', 'dish_soap', 'closetdrawer', 'pillow', 'bowl', 'dining_room', 'wall', 'powersocket', 'cup', 'food_cheese', 'clothes_gloves', 'coffe_maker', 'after_shave', 'kitchen_counter', 'dishrack', 'cupboard', 'sink']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Turn water on and run over dishes while scrubbing with a soapy sponge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'jelly', 'ground_coffee', 'nail_polish', 'detergent', 'orchid', 'needle', 'drawing', 'hanger', 'character', 'envelope', 'ceilinglamp', 'shower', 'electric_shaver', 'tooth_paste', 'mousepad', 'food_kiwi', 'powersocket', 'oven', 'tape', 'cup', 'photoframe', 'wooden_spoon', 'cards', 'kitchen_counter', 'cd', 'instrument_violin', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "grab the vesseles and wash", "actions": ["[WALK] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dining_room>", "[WASH] <bowl>", "[WALK] <detergent>", "[FIND] <detergent>", "[RINSE] <detergent>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'mop', 'dog', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'toaster', 'mat', 'closetdrawer', 'couch', 'food_jam', 'food_orange', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'food_cheese', 'keys', 'food_food', 'video_game_controller', 'television', 'ice', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Open fridge. Put groceries in fridge.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to home office, sit in the desk, switch on the computer and write a message", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mail>", "[OPEN] <mail>", "[TYPE] <keyboard>", "[CLOSE] <mail>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'band-aids', 'drawing', 'longboard', 'floor', 'ceiling', 'folder', 'computer', 'clothes_jacket', 'alcohol', 'ceilinglamp', 'balanceball', 'window', 'foundation', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'television', 'wine', 'towel_rack', 'pencil', 'shoes', 'dresser', 'cupboard', 'table', 'light', 'dishwasher']\nTask:Arrange furniture\nresponse:\n{\n    \"thoughts\": Move furniture where you want it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[PUSH] <coffee_table>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'blow_dryer', 'bathroom', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'stamp', 'keyboard', 'computer', 'food_bacon', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'coffe_maker', 'wooden_spoon', 'towel_rack', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'controller', 'light', 'soap']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug in vacuum. Walk around with vacuum, covering all of the floor. Unplug vacuum.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'juice', 'microwave', 'toilet', 'walllamp', 'orchid', 'bathroom_counter', 'bookshelf', 'bookmark', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'wall_clock', 'food_food', 'coffe_maker', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'cupboard', 'bathtub', 'home_office', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'milk', 'floor', 'desk', 'bedroom', 'food_cake', 'stovefan', 'tvstand', 'picture', 'window', 'clothes_socks', 'couch', 'microphone', 'oven', 'bench', 'toy', 'cutting_board', 'photoframe', 'coffe_maker', 'mouthwash', 'tray', 'pencil', 'controller', 'door', 'bathtub', 'home_office', 'light', 'bathroom_cabinet']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'alcohol', 'stovefan', 'mat', 'shoe-shine_kit', 'tvstand', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'wall', 'powersocket', 'photoframe', 'food_food', 'check', 'television', 'coffe_maker', 'ice', 'kitchen_counter', 'cupboard', 'hair', 'table', 'scrabble']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'juice', 'microwave', 'freezer', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'tea_bag', 'keyboard', 'computer', 'button', 'stovefan', 'mat', 'ceilinglamp', 'homework', 'closetdrawer', 'couch', 'wall', 'oven', 'cutting_board', 'toilet_paper', 'television', 'curtain', 'diary', 'pencil', 'shoes', 'dresser', 'table']\nTask:Put out flowers\nresponse:\n", "output": {"thoughts": "Place a flower centrepiece on the entrance hall table.", "actions": ["[WALK] <home_office>", "[WALK] <centerpiece>", "[FIND] <centerpiece>", "[GRAB] <centerpiece>", "[WALK] <table>", "[PUTBACK] <centerpiece> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on the TV, sit on the couch, find a show, and watch it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'blender', 'beer', 'mop', 'chair', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'ceiling', 'keyboard', 'alcohol', 'ceilinglamp', 'faucet', 'window', 'powersocket', 'toy', 'food_cheese', 'check', 'tray', 'console', 'mouse', 'pencil', 'kitchen_counter', 'hairbrush', 'hair', 'food_steak', 'bathtub', 'bathroom_cabinet', 'vase']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Wait until doorbell rings, look to see who it is outside and open the door if you feel so inclined.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'food_bread', 'toilet', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'longboard', 'shampoo', 'desk', 'stamp', 'keyboard', 'toaster', 'mat', 'deck_of_cards', 'faucet', 'foundation', 'mousepad', 'form', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'cupboard', 'door', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Travel through the house and locate the bathroom, and proceed to the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'dog', 'milk', 'floor', 'bedroom', 'computer', 'mail', 'alcohol', 'bookmark', 'mat', 'tvstand', 'window', 'couch', 'dustpan', 'pillow', 'dining_room', 'wall', 'wall_clock', 'coffee_table', 'novel', 'food_food', 'television', 'coin', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'maindoor', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit down in recliner.  Pick up novel off of coffee table.  Open novel to last read page.  Begin reading", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'chair', 'freezer', 'cpuscreen', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'stovefan', 'mousepad', 'wall', 'keys', 'check', 'oven_mitts', 'doorjamb', 'mouse', 'pencil', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'napkin', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": turn on the light for room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[SWITCHON] <light>', '[FIND] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'mail', 'mat', 'ceilinglamp', 'faucet', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'shelf', 'standingmirror', 'television', 'curtain', 'doorjamb', 'centerpiece', 'mouse', 'dresser', 'sink', 'table', 'clothes_shirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn TV on to TV channel and sit down on the couch and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'microwave', 'toilet', 'chair', 'walllamp', 'drawing', 'floor', 'ceiling', 'shampoo', 'computer', 'mat', 'food_orange', 'bowl', 'dining_room', 'wall', 'powersocket', 'oven', 'toy', 'shelf', 'cup', 'light_bulb', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'maindoor', 'bathtub', 'light', 'vase']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the kitchen, I find the light switch, and then i switch on the lights in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'walllamp', 'nail_polish', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'computer', 'mat', 'couch', 'clothes_pants', 'form', 'wall', 'toy', 'box', 'basket_for_clothes', 'light_bulb', 'novel', 'food_food', 'check', 'television', 'coffee_filter', 'toothbrush_holder', 'tray', 'laundry_detergent', 'clothes_dress', 'table', 'clothes_shirt', 'washing_machine']\nTask:Do laundry\nresponse:\n", "output": {"thoughts": "Collecting laundry from the children's hamper", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <basket_for_clothes> <table>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[POUR] <laundry_detergent> <washing_machine>", "[PUTOBJBACK] <laundry_detergent>", "[CLOSE] <filing_cabinet>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[PLUGIN] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'dog', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'alcohol', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'food_jam', 'pillow', 'wall', 'nightstand', 'cutting_board', 'novel', 'television', 'toothbrush_holder', 'mouse', 'door', 'dirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I pick up a book. I sit in a chair. I open the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['food_bread', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'detergent', 'nail_polish', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'piano_bench', 'shower', 'faucet', 'electric_shaver', 'closetdrawer', 'bowl', 'wall_clock', 'cutting_board', 'toilet_paper', 'check', 'tray', 'towel_rack', 'kettle', 'napkin', 'light', 'soap', 'creditcard']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Say hi", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'jelly', 'freezer', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'purse', 'computer', 'iron', 'window', 'mousepad', 'oil', 'wall', 'powersocket', 'bed', 'keys', 'mouthwash', 'wine', 'doorjamb', 'mouse', 'cupboard', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['jelly', 'toothbrush', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'shelf', 'standingmirror', 'keys', 'television', 'curtain', 'doorjamb', 'cards', 'napkin', 'home_office', 'table', 'light', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "We sit all around the television and watch a program.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'blender', 'food_salt', 'crayon', 'chair', 'band-aids', 'ceilingfan', 'needle', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'brush', 'ceilinglamp', 'shower', 'dining_room', 'wall', 'oven', 'powersocket', 'coffe_maker', 'curtain', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'maindoor', 'light']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Using soft brush, carefully dust both Indonesian wedding doll statues.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <brush>', '[GRAB] <brush>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'microwave', 'freezer', 'detergent', 'drawing', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'ceilinglamp', 'iron', 'faucet', 'window', 'homework', 'bowl', 'wall', 'wall_clock', 'bench', 'shelf', 'newspaper', 'mouthwash', 'food_donut', 'doorjamb', 'coffee_pot', 'door', 'sink', 'bathroom_cabinet', 'light', 'board_game']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'food_bacon', 'alcohol', 'ceilinglamp', 'pillow', 'wall', 'powersocket', 'bed', 'curtain', 'doorjamb', 'kettle', 'mouse', 'dresser', 'cupboard', 'door', 'maindoor', 'light', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I open the room door and press the light button on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_socks', 'closetdrawer', 'mousepad', 'clothes_pants', 'wall', 'powersocket', 'oven', 'cup', 'newspaper', 'food_food', 'check', 'coffe_maker', 'shaving_cream', 'mouse', 'pencil', 'shoes', 'dresser', 'hair', 'conditioner', 'light', 'clothes_shirt', 'shoe_rack']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "Open drawer. Pull out pants and put them on. Take a shirt out and put it on. Put on socks and shoes.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTON] <clothes_socks>", "[CLOSE] <dresser>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'filing_cabinet', 'toilet', 'detergent', 'food_butter', 'orchid', 'needle', 'bills', 'stereo', 'window', 'dining_room', 'bench', 'oven', 'bed', 'novel', 'check', 'coffee_filter', 'tray', 'wooden_spoon', 'centerpiece', 'headset', 'conditioner', 'food_steak', 'napkin', 'home_office', 'soap', 'scrabble', 'dirt', 'coffee']\nTask:File documents\nresponse:\n{\n    \"thoughts\": I file all of my paperwork (including bills, receipts, documents, device instructions, etc.) in my black paperwork case that has many different files in order to organize.  I place each piece of paper in its' corresponding subject (as indicated by printed words on each tab).,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bills>', '[FIND] <bills>', '[GRAB] <bills>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTIN] <bills> <filing_cabinet>', '[FIND] <check>', '[GRAB] <check>', '[WALK] <filing_cabinet>', '[PUTIN] <check> <filing_cabinet>', '[FIND] <check>', '[GRAB] <check>', '[WALK] <filing_cabinet>', '[PUTIN] <check> <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['juice', 'microwave', 'jelly', 'freezer', 'walllamp', 'food_carrot', 'dough', 'orchid', 'drawing', 'kitchen_cabinet', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'mat', 'plate', 'food_jam', 'wall', 'oven', 'bench', 'cutting_board', 'cup', 'food_vegetable', 'food_food', 'coffe_maker', 'ice', 'wine', 'tray', 'kitchen_counter', 'napkin', 'table']\nTask:Fix snack\nresponse:\n", "output": {"thoughts": "Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <kitchen_cabinet>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTBACK] <food_carrot> <plate>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[PUTBACK] <food_vegetable> <plate>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'juice', 'microwave', 'knifeblock', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'piano_bench', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'stovefan', 'tvstand', 'dish_soap', 'wall', 'pasta', 'coffee_filter', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'sink', 'scrabble']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <sink>', '[SWITCHOFF] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[GRAB] <plate>', '[SCRUB] <plate>', '[PUTOBJBACK] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[PUTOBJBACK] <plate>', '[SWITCHON] <faucet>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[GRAB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'filing_cabinet', 'tablelamp', 'crayon', 'woman', 'wallshelf', 'shampoo', 'floor', 'phone', 'desk', 'man', 'bookshelf', 'closetdrawer', 'couch', 'dining_room', 'nightstand', 'child', 'bed', 'photoframe', 'standingmirror', 'toilet_paper', 'oven_mitts', 'towel_rack', 'kitchen_counter', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'vase']\nTask:Say goodbye to guests leaving\nresponse:\n", "output": {"thoughts": "Tell them bye", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[FIND] <man>", "[TOUCH] <man>", "[FIND] <woman>", "[TOUCH] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'crayon', 'walllamp', 'woman', 'knifeblock', 'hanger', 'ceiling', 'folder', 'character', 'man', 'mop_bucket', 'bookmark', 'shower', 'window', 'electric_shaver', 'pillow', 'dining_room', 'wall_clock', 'cutting_board', 'bed', 'television', 'wine', 'curtain', 'mouse', 'pencil', 'sink', 'home_office', 'light', 'dirt', 'shoe_rack']\nTask:Greet people\nresponse:\n{\n    \"thoughts\": Say HI to those that come in.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <woman>', '[FIND] <woman>', '[GREET] <woman>', '[FIND] <woman>', '[GREET] <woman>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['sheets', 'freezer', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'envelope', 'computer', 'stereo', 'mousepad', 'closetdrawer', 'oil', 'clothes_pants', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'coin', 'rag', 'kettle', 'mouse', 'chessboard', 'dresser', 'light', 'clothes_shirt']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "Find dresser, grab pants, grab shirt, put on pants, put on shirt", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'pencil', 'bathtub', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": For browsing internet walk to home office. Find Computer. Walk to Computer. Find Chair. Sit in a chair and switch on computer. Connect with internet and browse the net.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'food_carrot', 'toothbrush', 'orchid', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'thread', 'stovefan', 'mat', 'toaster', 'shower', 'window', 'wall', 'powersocket', 'novel', 'food_food', 'check', 'television', 'rag', 'doorjamb', 'dresser', 'food_steak', 'spoon', 'bathroom_cabinet', 'light', 'table']\nTask:Study\nresponse:\n", "output": {"thoughts": "Do homework or read books while sitting on the chair.", "actions": ["[WALK] <bedroom>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'keys', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I want to relax on the sofa to some relaxing music in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <stereo>', '[PLUGIN] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'video_game_console', 'computer', 'button', 'mat', 'ceilinglamp', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'cutting_board', 'cup', 'standingmirror', 'video_game_controller', 'television', 'coffee_filter', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table', 'soap', 'dirt']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'knifeblock', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'faucet', 'couch', 'dustpan', 'pillow', 'towel', 'wall', 'nightstand', 'oven', 'box', 'bed', 'keys', 'check', 'toilet_paper', 'television', 'kitchen_counter', 'dresser', 'instrument_guitar', 'soap', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to living room.  Find sofa.  Sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_bread', 'microwave', 'freezer', 'walllamp', 'drawing', 'orchid', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'mat', 'shoe-shine_kit', 'wall', 'bench', 'oven', 'cutting_board', 'food_vegetable', 'food_food', 'coffe_maker', 'coffee_filter', 'wine', 'tray', 'food_onion', 'kitchen_counter', 'controller', 'food_steak', 'table']\nTask:Chop vegetables\nresponse:\n", "output": {"thoughts": "Cut up veggies with knife", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <cutting_board>", "[GRAB] <cutting_board>", "[CLOSE] <kitchen_cabinet>", "[WALK] <table>", "[PUTBACK] <knife> <table>", "[PUTBACK] <cutting_board> <table>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <food_vegetable> <cutting_board>", "[GRAB] <knife>", "[CUT] <food_vegetable>", "[PUTOBJBACK] <knife>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'freezer', 'detergent', 'trashcan', 'drawing', 'longboard', 'floor', 'phone', 'purse', 'computer', 'bookmark', 'brush', 'window', 'wall', 'child', 'powersocket', 'bed', 'novel', 'food_food', 'after_shave', 'curtain', 'doorjamb', 'towel_rack', 'shoes', 'laundry_detergent', 'door', 'napkin', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Read story to children,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <child>', '[GREET] <child>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[WALK] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'milk', 'knife', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'purse', 'computer', 'mat', 'tvstand', 'foundation', 'window', 'mousepad', 'closetdrawer', 'couch', 'bowl', 'wall', 'powersocket', 'keys', 'coffee_filter', 'mouse', 'dresser', 'home_office', 'light', 'coffee']\nTask:Hang keys\nresponse:\n", "output": {"thoughts": "Remove keys from purse, pocket or hand.  Place in designated area (bowl, or on hook). Leave area", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <purse>", "[GRAB] <purse>", "[OPEN] <purse>", "[FIND] <keys>", "[GRAB] <keys>", "[FIND] <bowl>", "[PUTBACK] <keys> <bowl>", "[CLOSE] <purse>", "[PUTOBJBACK] <purse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'walllamp', 'food_carrot', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'button', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'basket_for_clothes', 'food_food', 'television', 'curtain', 'dresser', 'clothes_dress', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Take TV remote to change TV Channel. Press channel/program +/- button on TV remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[TOUCH] <button>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'band-aids', 'freezer', 'bathroom', 'dough', 'bathroom_counter', 'floor', 'ceiling', 'mail', 'stovefan', 'mat', 'tvstand', 'faucet', 'towel', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'tray', 'cards', 'console', 'dresser', 'hair', 'sink', 'bathroom_cabinet', 'soap', 'instrument_violin', 'laser_pointer', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap and hot water", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'blender', 'crayon', 'walllamp', 'cat', 'drawing', 'orchid', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'bookmark', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'oven', 'cup', 'bed', 'clothes_gloves', 'newspaper', 'coffe_maker', 'curtain', 'hairbrush', 'food_steak', 'bag', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, find sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'measuring_cup', 'toilet', 'chair', 'drawing', 'food_egg', 'clothes_underwear', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'bookshelf', 'button', 'shoe-shine_kit', 'shower', 'window', 'broom', 'wall', 'wall_clock', 'powersocket', 'bench', 'standingmirror', 'novel', 'coin', 'doorjamb', 'door', 'light', 'table', 'creditcard']\nTask:Read\nresponse:\n", "output": {"thoughts": "We pick out a book from the bookshelf and read it.", "actions": ["[WALK] <novel>", "[WALK] <bookshelf>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'beer', 'freezer', 'walllamp', 'bathroom', 'bathroom_counter', 'character', 'bedroom', 'desk', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'clothes_pants', 'towel', 'bench', 'basket_for_clothes', 'bed', 'coin', 'food_donut', 'curtain', 'towel_rack', 'mouse', 'cd', 'comb', 'laundry_detergent', 'bathtub', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['microwave', 'bathroom', 'wallshelf', 'character', 'desk', 'mail', 'food_bacon', 'computer', 'mat', 'window', 'oil', 'closetdrawer', 'couch', 'dining_room', 'love_seat', 'cup', 'newspaper', 'standingmirror', 'coffe_maker', 'video_game_controller', 'television', 'centerpiece', 'cards', 'kitchen_counter', 'chessboard', 'maindoor', 'hairdryer', 'bathroom_cabinet', 'table', 'shoe_rack']\nTask:Read\nresponse:\n", "output": {"thoughts": "Reading a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'walllamp', 'drawing', 'needle', 'orchid', 'floor', 'stamp', 'keyboard', 'plate', 'faucet', 'shower', 'dish_soap', 'oil', 'wall', 'keys', 'coffe_maker', 'coffee_filter', 'pencil', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'controller', 'pajamas', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'bills', 'floor', 'computer', 'toaster', 'stovefan', 'shoe-shine_kit', 'faucet', 'ceilinglamp', 'dustpan', 'microphone', 'food_orange', 'dining_room', 'wall', 'cutting_board', 'toy', 'box', 'coffe_maker', 'food_donut', 'after_shave', 'mouse', 'kitchen_counter', 'shoes', 'hair', 'door', 'sink', 'light', 'soap', 'tooth_paste']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_salt', 'toilet', 'chair', 'bathroom', 'trashcan', 'drawing', 'floor', 'character', 'desk', 'purse', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'foundation', 'couch', 'pillow', 'nightstand', 'powersocket', 'novel', 'tray', 'mouse', 'bathroom_cabinet', 'soap', 'teeth']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'walllamp', 'needle', 'wallshelf', 'cpuscreen', 'hanger', 'bedroom', 'keyboard', 'bookmark', 'stovefan', 'shower', 'oil', 'dining_room', 'wall', 'powersocket', 'bench', 'love_seat', 'standingmirror', 'novel', 'coffe_maker', 'check', 'after_shave', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'maindoor', 'bathtub', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[READ] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[PUTBACK] <bookmark> <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'scissors', 'food_salt', 'toilet', 'chair', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'mail', 'toaster', 'homework', 'mousepad', 'clothes_pants', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'food_food', 'coffee_pot', 'mouse', 'home_office', 'table', 'dishwasher', 'vase', 'tooth_paste']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom. Pull down pants. Use the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <clothes_pants>', '[FIND] <clothes_pants>', '[PULL] <clothes_pants>', '[FIND] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'walllamp', 'hanger', 'floor', 'ceiling', 'keyboard', 'computer', 'stovefan', 'stereo', 'shower', 'window', 'mousepad', 'wall', 'nightstand', 'wall_clock', 'toy', 'photoframe', 'coffe_maker', 'curtain', 'lighter', 'towel_rack', 'kitchen_counter', 'controller', 'bathtub', 'bathroom_cabinet', 'table', 'instrument_violin', 'creditcard']\nTask:Open bathroom window\nresponse:\n", "output": {"thoughts": "Make sure bathroom window is unlocked. Lift up bathroom window so it is open.", "actions": ["[WALK] <bathroom>", "[WALK] <window>", "[FIND] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'freezer', 'walllamp', 'bathroom', 'cpuscreen', 'bathroom_counter', 'ceiling', 'keyboard', 'computer', 'food_bacon', 'table_cloth', 'ceilinglamp', 'faucet', 'picture', 'clothes_socks', 'towel', 'dining_room', 'wall', 'pasta', 'bed', 'centerpiece', 'kitchen_counter', 'door', 'bathtub', 'home_office', 'light', 'table', 'colander']\nTask:Clean screen\nresponse:\n{\n    \"thoughts\": Use a soft rag to wipe the screen until there are no marks on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <computer>', '[WIPE] <computer>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'fork', 'tablelamp', 'toilet', 'walllamp', 'orchid', 'character', 'alcohol', 'faucet', 'mousepad', 'dustpan', 'wall', 'nightstand', 'oven', 'cup', 'photoframe', 'coffe_maker', 'razor', 'after_shave', 'cards', 'dishrack', 'sink', 'spoon', 'bathroom_cabinet', 'clothes_skirt', 'vase', 'clothes_shirt', 'dirt']\nTask:Shut off alarm\nresponse:\n", "output": {"thoughts": "Shutting of the alarm clock", "actions": ["[WALK] <bedroom>", "[WALK] <alarm_clock>", "[FIND] <alarm_clock>", "[GRAB] <alarm_clock>", "[SWITCHOFF] <alarm_clock>", "[PUTOBJBACK] <alarm_clock>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'bathroom', 'trashcan', 'dough', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'faucet', 'window', 'food_noodles', 'mousepad', 'pillow', 'dining_room', 'wall', 'bed', 'food_food', 'razor', 'rag', 'console', 'mouse', 'towel_rack', 'kitchen_counter', 'door', 'sink', 'bathtub', 'bathroom_cabinet']\nTask:Clean sink\nresponse:\n{\n    \"thoughts\": Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[WIPE] <faucet>', '[WIPE] <sink>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bills', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'television', 'toothbrush_holder', 'wine', 'doorjamb', 'mouse', 'dresser', 'hairbrush', 'controller', 'clothes_dress', 'home_office', 'light', 'table']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Press the power button on the computer to turn it on", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'tablelamp', 'crayon', 'microwave', 'freezer', 'bathroom', 'trashcan', 'ceilingfan', 'knifeblock', 'clothes_underwear', 'longboard', 'shampoo', 'dvd_player', 'folder', 'mail', 'bookmark', 'toaster', 'shower', 'homework', 'couch', 'bed', 'clothes_gloves', 'towel_rack', 'mouse', 'pencil', 'cupboard', 'maindoor', 'table', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'jelly', 'chair', 'phone', 'ceiling', 'desk', 'bedroom', 'table_cloth', 'bookmark', 'stovefan', 'mat', 'tvstand', 'stereo', 'shoe-shine_kit', 'faucet', 'toaster', 'mousepad', 'powersocket', 'cutting_board', 'shelf', 'food_cheese', 'television', 'wooden_spoon', 'after_shave', 'food_onion', 'doorjamb', 'shoes', 'dishrack', 'vase', 'electrical_outlet']\nTask:Turn on radio\nresponse:\n", "output": {"thoughts": "Go to the radio.  Switch on the power button.  Tune the station.  Set the volume.  Listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <electrical_outlet>", "[TURNTO] <electrical_outlet>", "[PLUGIN] <stereo>", "[SWITCHON] <electrical_outlet>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'cat', 'food_carrot', 'toothbrush', 'drawing', 'orchid', 'floor', 'mail', 'stovefan', 'ceilinglamp', 'window', 'mousepad', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'photoframe', 'mouthwash', 'tray', 'curtain', 'coffee_pot', 'pajamas', 'door', 'sink', 'bathroom_cabinet', 'table', 'light', 'tooth_paste']\nTask:Relax\nresponse:\n{\n    \"thoughts\": Lie down on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'knife', 'laptop', 'tvstand', 'faucet', 'shower', 'window', 'mat', 'clothes_socks', 'closetdrawer', 'pillow', 'photoframe', 'standingmirror', 'newspaper', 'toothbrush_holder', 'lighter', 'centerpiece', 'kitchen_counter', 'shoes', 'cd', 'hair', 'conditioner', 'maindoor', 'bathtub', 'sink', 'table', 'vase']\nTask:Leave home\nresponse:\n", "output": {"thoughts": "Put on shoes. Put on coat. Open door and exit through it. Lock door behind you.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bookshelf', 'computer', 'stovefan', 'toaster', 'food_noodles', 'wall', 'oven', 'toy', 'food_vegetable', 'food_food', 'video_game_controller', 'water', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'dishwasher', 'scrabble']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": walk to kicthen , find stove , find cooking pot , find vegetables , find water  ,put vegetables in cooking pot, pour water to cooking pot,put cooking pot on stove, switch on stove,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <oven>', '[FIND] <pot>', '[FIND] <food_vegetable>', '[FIND] <water>', '[GRAB] <water>', '[WASH] <food_vegetable>', '[POUR] <water> <pot>', '[FIND] <pot>', '[GRAB] <pot>', '[PUTBACK] <pot> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'filing_cabinet', 'microwave', 'chair', 'bathroom', 'orchid', 'character', 'desk', 'bedroom', 'thread', 'computer', 'brush', 'ceilinglamp', 'faucet', 'stereo', 'closetdrawer', 'dining_room', 'wall', 'pasta', 'toy', 'cup', 'standingmirror', 'television', 'wine', 'doorjamb', 'kitchen_counter', 'conditioner', 'bathroom_cabinet', 'clothes_shirt']\nTask:Open door\nresponse:\n", "output": {"thoughts": "Open the door when leaving or coming in.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'tea_bag', 'mat', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'toy', 'photoframe', 'check', 'mouthwash', 'food_food', 'food_donut', 'wine', 'curtain', 'doorjamb', 'console', 'dresser', 'door', 'light']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['drawing', 'hanger', 'floor', 'ceiling', 'computer', 'brush', 'mat', 'ceilinglamp', 'faucet', 'stovefan', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cutting_board', 'shelf', 'toy', 'standingmirror', 'television', 'curtain', 'doorjamb', 'towel_rack', 'controller', 'home_office', 'table', 'soap', 'shoe_rack']\nTask:Turn off TV\nresponse:\n", "output": {"thoughts": "Walk over to tv and press power button", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to the living room, where my computer is. I locate it, switch it on and type desired URL adress in the web browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['food_bread', 'microwave', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'alcohol', 'piano_bench', 'stovefan', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'oil', 'food_peanut_butter', 'creditcard', 'form', 'wall', 'nightstand', 'oven', 'cutting_board', 'toy', 'newspaper', 'coffe_maker', 'tray', 'kitchen_counter', 'dresser', 'sink', 'electrical_outlet']\nTask:Toast bread\nresponse:\n", "output": {"thoughts": "Place two slices of bread in the toaster. Select medium toast on toaster.  Push down button to begin toasting.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[WALK] <electrical_outlet>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <electrical_outlet>", "[TURNTO] <toaster>", "[LOOKAT] <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'juice', 'scissors', 'beer', 'chair', 'freezer', 'bathroom', 'ceilingfan', 'hanger', 'floor', 'computer', 'food_bacon', 'mat', 'couch', 'creditcard', 'dining_room', 'oven', 'tape', 'bed', 'light_bulb', 'novel', 'food_food', 'mouse', 'cupboard', 'conditioner', 'home_office', 'table', 'scrabble', 'shoe_rack']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'ceilinglamp', 'laptop', 'faucet', 'mat', 'closetdrawer', 'teeth', 'dining_room', 'wall', 'pasta', 'toy', 'shelf', 'standingmirror', 'novel', 'check', 'television', 'curtain', 'console', 'dresser', 'napkin', 'table', 'light', 'scrabble', 'coffee']\nTask:Look at painting\nresponse:\n", "output": {"thoughts": "Walk to the painting.  Look up at the painting.  Admire it on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'photoframe', 'novel', 'television', 'mouse', 'kitchen_counter', 'door', 'sink', 'clothes_skirt', 'clothes_shirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to home office, find the book. Sit in the chair and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'measuring_cup', 'blow_dryer', 'ground_coffee', 'bathroom', 'cat', 'phone', 'ceiling', 'character', 'man', 'keyboard', 'computer', 'mat', 'window', 'mousepad', 'clothes_pants', 'powersocket', 'cutting_board', 'cup', 'standingmirror', 'bed', 'coffe_maker', 'television', 'food_rice', 'wine', 'controller', 'sink', 'napkin', 'laser_pointer', 'shoe_rack']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Walk up to the door.  Open the door and greet whomever is out side", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'beer', 'chair', 'bills', 'knife', 'hanger', 'bathroom_counter', 'floor', 'computer', 'stovefan', 'mat', 'tvstand', 'window', 'couch', 'pillow', 'towel', 'dining_room', 'food_kiwi', 'wall', 'toy', 'photoframe', 'novel', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'bag', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on couch. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <couch>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'beer', 'microwave', 'freezer', 'walllamp', 'dough', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'oven', 'cutting_board', 'shelf', 'coffe_maker', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'cards', 'kitchen_counter', 'maindoor', 'bathtub', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a can of beer, open it, sit on couch, relax", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <beer>", "[GRAB] <beer>", "[CLOSE] <freezer>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[DRINK] <beer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'walllamp', 'orchid', 'hanger', 'phone', 'computer', 'piano_bench', 'nightstand', 'pasta', 'wall_clock', 'powersocket', 'toy', 'food_cheese', 'clothes_gloves', 'check', 'toilet_paper', 'television', 'after_shave', 'doorjamb', 'door', 'hairdryer', 'sink', 'home_office', 'light', 'table', 'clothes_scarf', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": go to bedroom, sit down, pick up book, read,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sponge', 'scissors', 'tablelamp', 'measuring_cup', 'microwave', 'chair', 'band-aids', 'bathroom', 'milk', 'wallshelf', 'phone', 'character', 'computer', 'mail', 'button', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'standingmirror', 'centerpiece', 'kitchen_counter', 'shoes', 'maindoor', 'sink', 'bathtub', 'bathroom_cabinet', 'table']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Walk to the lamp.  Find the switch.  Click the switch off.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[SWITCHOFF] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'band-aids', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'homework', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'keys', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'light', 'soap']\nTask:Change TV channels\nresponse:\n{\n    \"thoughts\": Take remote. Change to desired channel by you.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'needle', 'floor', 'phone', 'mail', 'mat', 'plate', 'faucet', 'stereo', 'couch', 'dustpan', 'dining_room', 'wall', 'bench', 'oven', 'toy', 'cup', 'newspaper', 'standingmirror', 'food_food', 'razor', 'console', 'towel_rack', 'bathtub', 'table']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'toaster', 'plate', 'faucet', 'stereo', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'coffe_maker', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'bathtub', 'dishwasher', 'clothes_skirt', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Take a dish and put it in the dishwasher to be washed,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['blow_dryer', 'bathroom', 'food_carrot', 'toothbrush', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'clothes_jacket', 'mousepad', 'closetdrawer', 'clothes_pants', 'wall', 'powersocket', 'bench', 'cutting_board', 'bed', 'photoframe', 'television', 'food_onion', 'mouse', 'dresser', 'hairbrush', 'controller', 'light', 'clothes_shirt']\nTask:Organize closet\nresponse:\n", "output": {"thoughts": "Organize the hanging clothes in my closet by color and type and get rid of clothes I don't wear.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PULL] <clothes_pants>", "[PUTIN] <clothes_pants> <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PULL] <clothes_pants>", "[PUTIN] <clothes_pants> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUSH] <clothes_shirt>", "[PUTIN] <clothes_shirt> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUSH] <clothes_shirt>", "[PUTIN] <clothes_shirt> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[CLOSE] <dresser>", "[WALK] <basket_for_clothes>", "[PUTBACK] <clothes_shirt> <basket_for_clothes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'drawing', 'needle', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'dustpan', 'dining_room', 'wall', 'nightstand', 'powersocket', 'food_food', 'toilet_paper', 'doorjamb', 'coffee_pot', 'mouse', 'kettle', 'dresser', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk into living room, walk to light, turn on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'freezer', 'walllamp', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'mail', 'stovefan', 'laptop', 'window', 'towel', 'dining_room', 'food_kiwi', 'wall', 'oven', 'cutting_board', 'bed', 'coffe_maker', 'food_food', 'check', 'tray', 'kitchen_counter', 'dishrack', 'conditioner', 'food_steak', 'bathtub', 'table', 'vase']\nTask:Look in refrigerator\nresponse:\n", "output": {"thoughts": "Approach the refrigerator. Use the handle to open the refrigerator door. Look at each shelf of the refrigerator. Look in each refrigerator drawer. Close the refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[WALK] <freezer>", "[OPEN] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <food_food>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'detergent', 'milk', 'knifeblock', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'computer', 'stovefan', 'toaster', 'tvstand', 'faucet', 'iron', 'couch', 'wall', 'oven', 'food_food', 'television', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'laser_pointer', 'scrabble']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Cooking a meal,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[CLOSE] <kitchen_cabinet>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <pot> <sink>', '[GRAB] <pot>', '[SWITCHOFF] <faucet>', '[WALK] <oven>', '[PUTBACK] <pot> <oven>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[CLOSE] <freezer>', '[WALK] <oven>', '[PUTBACK] <food_food> <pot>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['pot', 'crayon', 'chair', 'bathroom', 'food_butter', 'phone', 'ceiling', 'bedroom', 'tea_bag', 'laptop', 'ceilinglamp', 'shower', 'iron', 'clothes_socks', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'powersocket', 'standingmirror', 'photoframe', 'toothbrush_holder', 'food_donut', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'controller', 'door', 'table']\nTask:Switch on lamp\nresponse:\n", "output": {"thoughts": "Press the button so the light comes on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'mop', 'microwave', 'dog', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'form', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'centerpiece', 'sink', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room. I bend down. I sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'window', 'clothes_pants', 'pillow', 'wall', 'wall_clock', 'oven', 'nightstand', 'cutting_board', 'shelf', 'clothes_gloves', 'food_food', 'coffe_maker', 'wooden_spoon', 'tray', 'kitchen_counter', 'sink', 'bathroom_cabinet']\nTask:Keep an eye on stove as something is cooking\nresponse:\n", "output": {"thoughts": "Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <pot>", "[TURNTO] <pot>", "[LOOKAT] <pot>", "[FIND] <wall_clock>", "[TURNTO] <wall_clock>", "[LOOKAT] <wall_clock>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'toaster', 'mat', 'ceilinglamp', 'plate', 'stovefan', 'dish_soap', 'broom', 'wall', 'bench', 'food_food', 'coffe_maker', 'tray', 'wooden_spoon', 'towel_rack', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'table', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go the in kitchen then grab the plate on the table. I put the plate and dish soap in the dishwasher. Finally, I turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'freezer', 'blow_dryer', 'walllamp', 'nail_polish', 'bathroom', 'toothbrush', 'drawing', 'hanger', 'floor', 'shampoo', 'character', 'desk', 'bookshelf', 'purse', 'envelope', 'mat', 'ceilinglamp', 'window', 'towel', 'wall', 'powersocket', 'toy', 'cup', 'food_food', 'coffe_maker', 'sink', 'duster']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Apply duster to shelves and clean them until dust is gone", "actions": ["[WALK] <home_office>", "[WALK] <duster>", "[FIND] <duster>", "[GRAB] <duster>", "[FIND] <filing_cabinet>", "[WIPE] <filing_cabinet>", "[PUTOBJBACK] <duster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'trashcan', 'hanger', 'floor', 'bedroom', 'computer', 'button', 'brush', 'mat', 'shower', 'faucet', 'window', 'clothes_hat', 'wall', 'child', 'oven', 'bed', 'newspaper', 'novel', 'curtain', 'centerpiece', 'doorjamb', 'mouse', 'cd', 'dresser', 'hairdryer', 'clothes_scarf', 'vase', 'board_game']\nTask:Story reading time\nresponse:\n{\n    \"thoughts\": Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[TURNTO] <child>', '[LOOKAT] <child>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'glue', 'blender', 'food_salt', 'character', 'bedroom', 'mat', 'tvstand', 'food_noodles', 'garbage_can', 'couch', 'nightstand', 'wall_clock', 'bench', 'powersocket', 'toy', 'cutting_board', 'tape', 'food_vegetable', 'food_food', 'food_donut', 'tray', 'mouse', 'pencil', 'door', 'bathtub', 'home_office', 'table', 'colander', 'pantry']\nTask:Organize pantry\nresponse:\n", "output": {"thoughts": "Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.", "actions": ["[WALK] <dining_room>", "[WALK] <pantry>", "[FIND] <food_food>", "[GRAB] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[FIND] <garbage_can>", "[PUTBACK] <food_food> <garbage_can>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <dry_pasta>", "[GRAB] <dry_pasta>", "[TURNTO] <dry_pasta>", "[LOOKAT] <dry_pasta>", "[PUTOBJBACK] <dry_pasta>", "[FIND] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[PUSH] <food_food>", "[FIND] <food_noodles>", "[GRAB] <food_noodles>", "[TURNTO] <food_noodles>", "[LOOKAT] <food_noodles>", "[PUTOBJBACK] <food_noodles>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'freezer', 'walllamp', 'milk', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'alcohol', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'mousepad', 'wall', 'oven', 'powersocket', 'cup', 'ice', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Get something to drink\nresponse:\n{\n    \"thoughts\": Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[CLOSE] <freezer>', '[POUR] <milk> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'dog', 'chair', 'dough', 'wallshelf', 'hanger', 'cleaning_solution', 'keyboard', 'computer', 'ceilinglamp', 'plate', 'shower', 'faucet', 'window', 'homework', 'clothes_hat', 'nightstand', 'wall_clock', 'cup', 'photoframe', 'novel', 'check', 'oven_mitts', 'curtain', 'rag', 'towel_rack', 'shoes', 'hairbrush', 'maindoor', 'spoon']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'milk', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'shower', 'window', 'couch', 'clothes_hat', 'pillow', 'wall', 'photoframe', 'check', 'mouthwash', 'razor', 'television', 'coin', 'coffe_maker', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room and I lie down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'jelly', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'laptop', 'stovefan', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'table']\nTask:Use laptop\nresponse:\n", "output": {"thoughts": "Sit down on a couch and perform tasks such as navigating the internet on a laptop", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <keyboard>", "[LOOKAT] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'bathroom', 'trashcan', 'milk', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'clothes_jacket', 'toaster', 'stovefan', 'tvstand', 'picture', 'window', 'dustpan', 'pillow', 'dining_room', 'wall', 'box', 'bed', 'wine', 'curtain', 'kettle', 'home_office', 'light', 'dishwasher']\nTask:Look out window\nresponse:\n{\n    \"thoughts\": I head into my child's room and look out the window to view the neighborhood from a different area in the house.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <window>', '[TURNTO] <window>', '[LOOKAT] <window>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'dough', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'oil', 'wall', 'powersocket', 'shelf', 'toilet_paper', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'light', 'table', 'vase']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Keep my keyboard vacuumed and free of debris.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'bed', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": Type on a computer in the home office.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'dog', 'walllamp', 'detergent', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'towel', 'wall', 'wall_clock', 'powersocket', 'box', 'television', 'doorjamb', 'centerpiece', 'mouse', 'dresser', 'door', 'bathtub', 'light', 'table']\nTask:Clean screen\nresponse:\n", "output": {"thoughts": "Use a soft rag to wipe the screen until there are no marks on it.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <computer>", "[WIPE] <computer>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'toilet', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'button', 'ceilinglamp', 'faucet', 'oil', 'teeth', 'clothes_hat', 'towel', 'wall', 'photoframe', 'bed', 'doorjamb', 'diary', 'kitchen_counter', 'dishrack', 'coffee', 'hairbrush', 'maindoor', 'sink', 'bathroom_cabinet', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom. Turn on the sink. Rinse hands. Apply soap and scrub your hands together. Rinse hands again. Turn off faucet. Dry hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'filing_cabinet', 'food_bread', 'microwave', 'walllamp', 'dough', 'woman', 'toothbrush', 'wallshelf', 'ceiling', 'character', 'man', 'toaster', 'stereo', 'window', 'couch', 'dustpan', 'pillow', 'nightstand', 'bench', 'newspaper', 'keys', 'standingmirror', 'television', 'curtain', 'sink', 'instrument_guitar', 'table', 'light']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Stand next to the door. When the door opens from the outside, greet the guests.", "actions": ["[RUN] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'laptop', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dishrack', 'door', 'light', 'shoe_rack']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office, walk to desk, sit in chair, turn on laptop, grab mouse, open a web page to look at it, move mouse, keep looking at laptop screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <mouse>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['food_butter', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'toaster', 'stereo', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'bench', 'cutting_board', 'cup', 'check', 'food_donut', 'food_rice', 'curtain', 'centerpiece', 'mouse', 'vacuum_cleaner', 'shoes', 'dresser', 'light', 'teeth']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Run the vacuum cleaner across the floor, back and forth a couple of times until clean.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'remote_control', 'alcohol', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'couch', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'curtain', 'rag', 'doorjamb', 'dresser', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": go to living room, sit on couch, get remote, switch on tv, watch tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'juice', 'freezer', 'walllamp', 'drawing', 'orchid', 'phone', 'ceiling', 'character', 'cleaning_solution', 'button', 'toaster', 'ceilinglamp', 'homework', 'oil', 'mousepad', 'pillow', 'dining_room', 'food_kiwi', 'pasta', 'oven', 'coffe_maker', 'food_food', 'television', 'rag', 'towel_rack', 'bathtub', 'instrument_guitar', 'vase']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Spray counter tops with all purpose cleaner. Using towel, wipe counter, removing all dirt and grime until clean.", "actions": ["[WALK] <dining_room>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'beer', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'foundation', 'window', 'wall', 'powersocket', 'cup', 'photoframe', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk into the Living room.  Find the television remote.  Use it to turn on the television.  Point the remote at the television. Press the channel button on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'chair', 'blow_dryer', 'nail_polish', 'trashcan', 'dough', 'needle', 'cpuscreen', 'stamp', 'thread', 'stereo', 'floor_lamp', 'mousepad', 'form', 'pillow', 'dining_room', 'nightstand', 'oven', 'bench', 'cup', 'standingmirror', 'coffe_maker', 'food_donut', 'kettle', 'door', 'home_office', 'light', 'vase']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Clean lamp", "actions": ["[WALK] <home_office>", "[WALK] <paper_towel>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WALK] <floor_lamp>", "[SWITCHOFF] <floor_lamp>", "[WIPE] <floor_lamp>", "[FIND] <trashcan>", "[PUTBACK] <paper_towel> <trashcan>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'glue', 'toilet', 'chair', 'detergent', 'dough', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'keyboard', 'alcohol', 'stovefan', 'ceilinglamp', 'couch', 'microphone', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'doorjamb', 'kitchen_counter', 'hairbrush', 'cupboard', 'home_office', 'dishwasher', 'coffee', 'shoe_rack']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'cat', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'food_cake', 'computer', 'brush', 'electric_shaver', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'toy', 'food_food', 'towel_rack', 'mouse', 'dresser', 'pajamas', 'sink', 'napkin', 'light', 'colander', 'tooth_paste']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTBACK] <toy> <dresser>", "[CLOSE] <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'walllamp', 'toothbrush', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'bench', 'box', 'cup', 'photoframe', 'novel', 'mouthwash', 'television', 'curtain', 'doorjamb', 'comb', 'light', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit and read,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[WALK] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'band-aids', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'phone', 'cleaning_solution', 'bookshelf', 'computer', 'stereo', 'oil', 'mousepad', 'wall', 'wall_clock', 'nightstand', 'shelf', 'newspaper', 'food_food', 'video_game_controller', 'coin', 'rag', 'mouse', 'pencil', 'bathtub', 'bathroom_cabinet', 'light', 'coffee', 'creditcard']\nTask:Clean  mirror\nresponse:\n", "output": {"thoughts": "GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'bathroom', 'ceilingfan', 'knife', 'drawing', 'needle', 'knifeblock', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'phone', 'piano_bench', 'faucet', 'window', 'teeth', 'clothes_pants', 'wall', 'photoframe', 'check', 'oven_mitts', 'tray', 'pencil', 'clothes_dress', 'sink', 'bathroom_cabinet', 'home_office', 'clothes_skirt', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[PUSH] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'milk', 'drawing', 'cpuscreen', 'shampoo', 'folder', 'bookshelf', 'computer', 'table_cloth', 'ceilinglamp', 'mat', 'oil', 'mousepad', 'couch', 'towel', 'nightstand', 'wall', 'cutting_board', 'cup', 'photoframe', 'bed', 'food_donut', 'mouse', 'kitchen_counter', 'pajamas', 'sink', 'napkin', 'home_office', 'tooth_paste']\nTask:Open front door\nresponse:\n", "output": {"thoughts": "Walk over to the front door, turn the knob, and open the door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'fork', 'filing_cabinet', 'tablelamp', 'chair', 'band-aids', 'floor', 'ceiling', 'stamp', 'alcohol', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'towel', 'wall', 'toy', 'bed', 'keys', 'photoframe', 'coffe_maker', 'television', 'coin', 'kitchen_counter', 'dresser', 'cupboard', 'food_steak', 'sink', 'table']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Run warm water in the sink. Pour dish washing liquid on the dishes. Pick up dishes and wash them individually with a sponge. Rinse suds off under running water. Dry dishes with towel.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[WASH] <plate>', '[RINSE] <plate>', '[RINSE] <sponge>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <sponge>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <plate>', '[PUTOBJBACK] <towel>', '[WALK] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'floor', 'ceiling', 'bedroom', 'window', 'clothes_socks', 'couch', 'dustpan', 'clothes_pants', 'food_orange', 'food_kiwi', 'wall', 'toy', 'bed', 'photoframe', 'food_food', 'coffe_maker', 'wine', 'doorjamb', 'diary', 'console', 'mouse', 'door', 'bathroom_cabinet', 'light', 'vase', 'soap']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Grasp light switch, flip light switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'food_kiwi', 'wall', 'powersocket', 'clothes_gloves', 'food_food', 'coffe_maker', 'doorjamb', 'coffee_pot', 'headset', 'door', 'hairdryer', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk towards to the bedroom, and I open the door of the bedroom, and I switch on the lights inside the bedroom, the switch is beside the door.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'walllamp', 'orchid', 'needle', 'clothes_underwear', 'dvd_player', 'character', 'desk', 'mop_bucket', 'table_cloth', 'brush', 'ceilinglamp', 'mat', 'stereo', 'dining_room', 'nightstand', 'powersocket', 'toy', 'shelf', 'check', 'toothbrush_holder', 'mouse', 'chessboard', 'dresser', 'laundry_detergent', 'home_office', 'bathroom_cabinet', 'teeth', 'creditcard']\nTask:Put toys away\nresponse:\n", "output": {"thoughts": "Pick up all the toys on the floor and put them in their correct storage bin or shelf.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'video_game_console', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'dustpan', 'pillow', 'wall', 'cup', 'video_game_controller', 'television', 'curtain', 'dishrack', 'napkin', 'table', 'vase', 'scrabble', 'shoe_rack']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on game.  Use controller,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[PLUGIN] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'crayon', 'tea', 'freezer', 'walllamp', 'bathroom', 'detergent', 'blow_dryer', 'drawing', 'floor', 'desk', 'thread', 'mop_bucket', 'tvstand', 'towel', 'wall', 'toy', 'cup', 'check', 'razor', 'food_donut', 'wine', 'centerpiece', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'laundry_detergent', 'light', 'soap']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Robot has a vacuum underneath its feat/body that will allow it to clean the surface of the floor.", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[WALK] <dining_room>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[WALK] <home_office>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <filing_cabinet>", "[PUTBACK] <vacuum_cleaner> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'sheets', 'food_bread', 'food_pizza', 'tablelamp', 'scissors', 'toilet', 'walllamp', 'woman', 'knifeblock', 'hanger', 'bathroom_counter', 'man', 'tea_bag', 'keyboard', 'mail', 'stereo', 'window', 'couch', 'child', 'powersocket', 'photoframe', 'novel', 'coffe_maker', 'video_game_controller', 'ice', 'rag', 'headset', 'chessboard', 'maindoor']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Greet guests when they knock on the door and invite them to enter my home,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'ground_coffee', 'cpuscreen', 'floor', 'ceiling', 'keyboard', 'toaster', 'ceilinglamp', 'mat', 'faucet', 'food_orange', 'bowl', 'wall', 'wall_clock', 'cutting_board', 'bed', 'photoframe', 'coffe_maker', 'mouthwash', 'coffee_filter', 'water', 'doorjamb', 'pencil', 'kitchen_counter', 'hair', 'controller', 'sink', 'soap']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[GRAB] <ground_coffee>", "[PUTOBJBACK] <ground_coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'keyboard', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'bench', 'bed', 'check', 'television', 'tray', 'curtain', 'mouse', 'cupboard', 'door', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'toilet', 'freezer', 'walllamp', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'phone', 'character', 'mail', 'food_bacon', 'stovefan', 'mat', 'mousepad', 'couch', 'dining_room', 'broom', 'wall', 'oven', 'cutting_board', 'coffe_maker', 'food_food', 'tray', 'cards', 'mouse', 'kitchen_counter', 'home_office', 'soap']\nTask:Put away groceries\nresponse:\n", "output": {"thoughts": "Place groceries in refrigerator from shopping at the grocery store.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'walllamp', 'trashcan', 'drawing', 'floor', 'ceiling', 'button', 'toaster', 'window', 'mousepad', 'clothes_pants', 'wall', 'basket_for_clothes', 'coffe_maker', 'check', 'television', 'toothbrush_holder', 'ice', 'diary', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'pajamas', 'napkin', 'coffee_cup', 'dishwasher', 'soap', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to the kitchen at the coffee maker, grab coffee, pour coffee into coffee maker, put coffee into coffee maker, push coffee maker's button, wait to complete, grab coffee cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[OPEN] <coffe_maker>', '[POUR] <coffee> <coffe_maker>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['sponge', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'oven', 'bench', 'shelf', 'check', 'television', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'light', 'table', 'vase']\nTask:Research\nresponse:\n", "output": {"thoughts": "I read different articles online to gain a better understanding of a particular business, industry, method, terminology, etc.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'drawing', 'floor', 'bookshelf', 'toaster', 'ceilinglamp', 'stereo', 'picture', 'window', 'couch', 'towel', 'wall', 'nightstand', 'oven', 'cup', 'television', 'tray', 'kitchen_counter', 'cd', 'dresser', 'hairbrush', 'cupboard', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to living room. Walk to radio. Switch on the radio. Walk to sofa and sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['toilet', 'band-aids', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'desk', 'keyboard', 'mail', 'button', 'table_cloth', 'bookmark', 'electric_shaver', 'pillow', 'food_kiwi', 'nightstand', 'shelf', 'food_cheese', 'light_bulb', 'check', 'coffe_maker', 'razor', 'mirror', 'food_food', 'centerpiece', 'kitchen_counter', 'sink', 'bathtub', 'home_office', 'table']\nTask:Shave\nresponse:\n", "output": {"thoughts": "Pick up razor.  Push the button to turn it on.  Move it in a linear motion against the beard growth.  Make sure to get the bottom of your chin, your cheeks, and mustache area.", "actions": ["[WALK] <bathroom>", "[WALK] <electric_shaver>", "[FIND] <electric_shaver>", "[GRAB] <electric_shaver>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[SWITCHON] <electric_shaver>", "[SWITCHOFF] <electric_shaver>", "[PUTOBJBACK] <electric_shaver>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'chair', 'bathroom', 'food_carrot', 'knifeblock', 'floor', 'phone', 'bedroom', 'envelope', 'mail', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'tvstand', 'shower', 'window', 'candle', 'microphone', 'pillow', 'basket_for_clothes', 'toilet_paper', 'check', 'rag', 'chessboard', 'laundry_detergent', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Pick up my dirty clothes, load the washer, add detergent, turn on washer,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[RUN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'stereo', 'shower', 'mousepad', 'closetdrawer', 'dustpan', 'wall', 'powersocket', 'box', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'light', 'table', 'printing_paper']\nTask:Print out papers\nresponse:\n", "output": {"thoughts": "Turn on printer and fill with blank paper. Use computer to click on printing option. Take printed page out of print.", "actions": ["[WALK] <home_office>", "[WALK] <fax_machine>", "[FIND] <fax_machine>", "[SWITCHON] <fax_machine>", "[FIND] <printing_paper>", "[GRAB] <printing_paper>", "[PUTBACK] <printing_paper> <fax_machine>", "[FIND] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[GRAB] <printing_paper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'pot', 'toilet', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'kettle', 'shaving_cream', 'kitchen_counter', 'chessboard', 'dresser', 'controller', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I get home from work and walk into the living room. I see the couch and go to sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['ironing_board', 'sheets', 'glue', 'microwave', 'toilet', 'walllamp', 'blow_dryer', 'bathroom_counter', 'bedroom', 'computer', 'stovefan', 'tvstand', 'iron', 'faucet', 'mousepad', 'clothes_socks', 'couch', 'toy', 'cup', 'keys', 'coffe_maker', 'after_shave', 'towel_rack', 'mouse', 'kitchen_counter', 'controller', 'maindoor', 'bathtub', 'tooth_paste', 'clothes_shirt']\nTask:Iron shirt\nresponse:\n", "output": {"thoughts": "Set up an ironing board and an iron.  Lay a shirt on board.  Run iron over shirt until it is free of wrinkles.", "actions": ["[WALK] <bedroom>", "[WALK] <ironing_board>", "[FIND] <ironing_board>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <ironing_board>", "[FIND] <iron>", "[GRAB] <iron>", "[PLUGIN] <iron>", "[PLUGOUT] <iron>", "[PUTOBJBACK] <iron>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'beer', 'crayon', 'detergent', 'drawing', 'food_egg', 'floor', 'ceiling', 'bedroom', 'brush', 'mat', 'ceilinglamp', 'shower', 'balanceball', 'window', 'couch', 'towel', 'wall', 'bench', 'light_bulb', 'television', 'water', 'curtain', 'mouse', 'kitchen_counter', 'door', 'home_office', 'soap', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>', '[TURNTO] <shower>', '[FIND] <water>', '[RINSE] <water>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <water>', '[FIND] <towel>', '[GRAB] <towel>', '[SCRUB] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'cpuscreen', 'floor', 'toaster', 'mat', 'iron', 'window', 'electric_shaver', 'homework', 'clothes_socks', 'couch', 'candle', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'standingmirror', 'razor', 'television', 'tray', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'comb', 'door', 'table']\nTask:Pull up carpet\nresponse:\n", "output": {"thoughts": "Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.", "actions": ["[WALK] <home_office>", "[WALK] <mat>", "[FIND] <mat>", "[TOUCH] <mat>", "[PULL] <mat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'chair', 'trashcan', 'drawing', 'clothes_underwear', 'floor', 'phone', 'character', 'keyboard', 'mat', 'ceilinglamp', 'plate', 'tvstand', 'shower', 'faucet', 'mousepad', 'couch', 'wall', 'bench', 'powersocket', 'newspaper', 'bed', 'television', 'mouse', 'controller', 'table', 'light', 'instrument_violin', 'creditcard']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.,\n    \"actions\": ['[WALK] <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_carrot', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'mail', 'stovefan', 'mat', 'faucet', 'mousepad', 'towel', 'pillow', 'wall', 'wall_clock', 'oven', 'toy', 'cup', 'hands_both', 'standingmirror', 'food_donut', 'tray', 'mouse', 'kitchen_counter', 'hairdryer', 'sink', 'bathroom_cabinet', 'soap', 'teeth', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Turn faucet clockwise to turn on water.  Place hands under stream of water, pick up soap, rub hands with soap.  Put soap back down.  Rub hands together with soap on them.  Rinse hands under water from faucet.  Turn faucet counterclockwise to turn water off.  Dry hands by rubbing them on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'detergent', 'trashcan', 'floor', 'ceiling', 'bedroom', 'toaster', 'mat', 'stereo', 'pillow', 'wall', 'cup', 'bed', 'coffe_maker', 'food_food', 'coffee_filter', 'water', 'food_donut', 'wooden_spoon', 'food_onion', 'doorjamb', 'pencil', 'kitchen_counter', 'cupboard', 'napkin', 'spoon', 'dishwasher', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": I find the coffee machine, fill it with water and coffee then press the button to make the coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[WALK] <coffee>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'food_butter', 'knife', 'orchid', 'floor', 'character', 'desk', 'food_bacon', 'table_cloth', 'piano_bench', 'mat', 'picture', 'window', 'couch', 'pillow', 'bowl', 'wall', 'nightstand', 'child', 'light_bulb', 'toilet_paper', 'television', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'bathtub', 'table', 'board_game']\nTask:Let baby learn how to walk\nresponse:\n", "output": {"thoughts": "The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[GREET] <child>", "[FIND] <couch>", "[TURNTO] <couch>", "[POINTAT] <couch>", "[TURNTO] <child>", "[LOOKAT] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'mop', 'toilet', 'chair', 'trashcan', 'knifeblock', 'hanger', 'longboard', 'floor', 'phone', 'keyboard', 'balanceball', 'foundation', 'pillow', 'wall', 'powersocket', 'box', 'photoframe', 'bed', 'check', 'rag', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'maindoor', 'light', 'colander']\nTask:Sit\nresponse:\n{\n    \"thoughts\": We sit at the dining room table,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['scissors', 'microwave', 'toilet', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'mop_bucket', 'stovefan', 'mat', 'faucet', 'mousepad', 'towel', 'wall', 'wall_clock', 'oven', 'bench', 'cutting_board', 'shelf', 'food_rice', 'tray', 'after_shave', 'towel_rack', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'soap', 'laser_pointer', 'dirt', 'hands_both']\nTask:Hand washing\nresponse:\n", "output": {"thoughts": "Turn on the faucet. Soap up your hands. Then rinse under the running water of the faucet. Turn off faucet and dry hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'blow_dryer', 'cat', 'wallshelf', 'ceiling', 'desk', 'bedroom', 'keyboard', 'stovefan', 'mousepad', 'powersocket', 'oven', 'bed', 'food_food', 'check', 'wooden_spoon', 'towel_rack', 'mouse', 'dishrack', 'comb', 'maindoor', 'bathtub', 'napkin', 'home_office', 'light', 'table', 'soap', 'laser_pointer', 'scrabble']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <address_book>', '[FIND] <address_book>', '[GRAB] <address_book>', '[READ] <address_book>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'detergent', 'food_carrot', 'orchid', 'floor', 'wine_glass', 'food_cake', 'mat', 'picture', 'window', 'mousepad', 'couch', 'clothes_hat', 'pillow', 'wall', 'pasta', 'shelf', 'food_cheese', 'newspaper', 'food_food', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'bathtub', 'table', 'vase']\nTask:Entertain\nresponse:\n", "output": {"thoughts": "Give guests wine, cheese, crackers and sit and talk", "actions": ["[WALK] <home_office>", "[WALK] <food_cheese>", "[FIND] <food_cheese>", "[FIND] <food_food>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <wine_glass>", "[GRAB] <wine_glass>", "[DRINK] <wine_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'drawing', 'needle', 'knife', 'clothes_underwear', 'hanger', 'floor', 'envelope', 'computer', 'ceilinglamp', 'oil', 'mousepad', 'pillow', 'bowl', 'wall', 'nightstand', 'child', 'bed', 'novel', 'food_food', 'coffe_maker', 'television', 'doorjamb', 'dishrack', 'comb', 'laundry_detergent', 'conditioner', 'sink', 'instrument_violin']\nTask:Story reading time\nresponse:\n{\n    \"thoughts\": Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[TURNTO] <child>', '[LOOKAT] <child>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['blender', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'clothes_jacket', 'table_cloth', 'mat', 'floor_lamp', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'light_bulb', 'novel', 'mouthwash', 'television', 'ice', 'curtain', 'after_shave', 'doorjamb', 'console', 'kettle', 'conditioner', 'bathtub', 'sink', 'table', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'trashcan', 'orchid', 'knifeblock', 'floor', 'tea_bag', 'keyboard', 'computer', 'stovefan', 'tvstand', 'stereo', 'window', 'dustpan', 'candle', 'food_jam', 'wall', 'pasta', 'bed', 'photoframe', 'coffe_maker', 'television', 'tray', 'curtain', 'wooden_spoon', 'towel_rack', 'pencil', 'chessboard', 'cupboard', 'maindoor', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Put iPhone into speaker holder, open up the Pandora app, listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[WALK] <stereo>', '[PLUGIN] <phone>', '[SWITCHON] <stereo>', '[FIND] <bed>', '[SIT] <bed>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'food_ice_cream', 'beer', 'microwave', 'toilet', 'freezer', 'walllamp', 'food_egg', 'floor', 'ceiling', 'phone', 'keyboard', 'toaster', 'stovefan', 'brush', 'faucet', 'food_peanut_butter', 'clothes_pants', 'wall', 'oven', 'powersocket', 'cutting_board', 'food_cheese', 'coffe_maker', 'food_food', 'tray', 'kitchen_counter', 'comb', 'sink']\nTask:Make toast\nresponse:\n", "output": {"thoughts": "Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'measuring_cup', 'microwave', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'face', 'mop_bucket', 'tvstand', 'faucet', 'food_peanut_butter', 'towel', 'wall', 'oven', 'face_soap', 'tape', 'photoframe', 'bed', 'keys', 'rag', 'cards', 'comb', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet']\nTask:Wash face\nresponse:\n{\n    \"thoughts\": Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <rag>', '[GRAB] <rag>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[WASH] <rag>', '[FIND] <face_soap>', '[GRAB] <face_soap>', '[POUR] <face_soap> <rag>', '[PUTOBJBACK] <face_soap>', '[FIND] <face>', '[SCRUB] <face>', '[PUTBACK] <rag> <sink>', '[RINSE] <face>', '[RINSE] <rag>', '[SWITCHOFF] <faucet>', '[GRAB] <rag>', '[PUTOBJBACK] <rag>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <face>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'legs_both', 'desk', 'keyboard', 'face', 'computer', 'toaster', 'mousepad', 'closetdrawer', 'dustpan', 'towel', 'wall', 'feet_both', 'powersocket', 'arms_both', 'basket_for_clothes', 'cup', 'check', 'mouse', 'shoes', 'dresser', 'hairbrush', 'controller', 'light', 'clothes_shirt']\nTask:Try yourself off\nresponse:\n", "output": {"thoughts": "After your shower, go into your bedroom and use a towel to dry water off of yourself before dressing.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <towel>", "[GRAB] <towel>", "[CLOSE] <dresser>", "[FIND] <face>", "[WIPE] <face>", "[FIND] <arms_both>", "[WIPE] <arms_both>", "[FIND] <legs_both>", "[WIPE] <legs_both>", "[FIND] <feet_both>", "[WIPE] <feet_both>", "[FIND] <basket_for_clothes>", "[PUTBACK] <towel> <basket_for_clothes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'ground_coffee', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'folder', 'purse', 'computer', 'toaster', 'laptop', 'shower', 'picture', 'couch', 'wall', 'pasta', 'nightstand', 'box', 'photoframe', 'coffe_maker', 'television', 'coffee_filter', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'instrument_violin']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open the coffee maker and put a filer, grab ground coffee and put it in the coffee maker, then close the coffee maker and switch it on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[FIND] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'tea', 'walllamp', 'bathroom', 'wallshelf', 'ceiling', 'dvd_player', 'tea_bag', 'food_bacon', 'toaster', 'homework', 'clothes_pants', 'nightstand', 'wall', 'powersocket', 'toy', 'newspaper', 'standingmirror', 'novel', 'bed', 'console', 'mouse', 'shoes', 'controller', 'door', 'sink', 'bathtub', 'home_office', 'table', 'vase']\nTask:Watch fly\nresponse:\n", "output": {"thoughts": "Look around for a fly, often on a wall or ceiling.  Follow it's movements.", "actions": ["[WALK] <bedroom>", "[WALK] <fly>", "[FIND] <fly>", "[TURNTO] <fly>", "[LOOKAT] <fly>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'foundation', 'couch', 'wall', 'pasta', 'toy', 'box', 'television', 'curtain', 'diary', 'towel_rack', 'kitchen_counter', 'cupboard', 'clothes_scarf', 'home_office', 'table', 'vase', 'coffee']\nTask:Change TV channels\nresponse:\n{\n    \"thoughts\": Take remote. Change to desired channel by you.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'juice', 'nail_polish', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'mat', 'shower', 'window', 'closetdrawer', 'wall', 'nightstand', 'cutting_board', 'bed', 'photoframe', 'coffe_maker', 'television', 'curtain', 'mouse', 'pencil', 'vacuum_cleaner', 'hair', 'maindoor', 'table', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I lie on my bed and read a book.", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'dresser', 'hairbrush', 'clothes_dress', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'detergent', 'wallshelf', 'hanger', 'floor', 'phone', 'shampoo', 'character', 'desk', 'mop_bucket', 'bookmark', 'toaster', 'mat', 'window', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'mouthwash', 'television', 'razor', 'curtain', 'after_shave', 'doorjamb', 'bathroom_cabinet', 'table', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up a book and open to the last page you were on. Read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'cat', 'ceilingfan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'oil', 'couch', 'wall', 'wall_clock', 'powersocket', 'oven', 'bench', 'newspaper', 'mouthwash', 'television', 'diary', 'towel_rack', 'mouse', 'dresser', 'hairbrush', 'hairdryer', 'soap']\nTask:Check email\nresponse:\n{\n    \"thoughts\": Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[WALK] <chair>', '[SIT] <chair>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <laptop>']\n}\n", "input": "\nScene_Objects:['microwave', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'plate', 'iron', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'cutting_board', 'clothes_gloves', 'check', 'coffe_maker', 'food_food', 'curtain', 'food_onion', 'mouse', 'comb', 'dresser', 'clothes_dress', 'sink', 'light', 'coffee']\nTask:Hang up jacket\nresponse:\n", "output": {"thoughts": "I put clothes on hangers and hang them in the closet", "actions": ["[WALK] <bedroom>", "[WALK] <clothes_dress>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[WALK] <dresser>", "[FIND] <hanger>", "[OPEN] <dresser>", "[PUTIN] <clothes_dress> <dresser>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[WALK] <dresser>", "[PUTIN] <clothes_dress> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'juice', 'jelly', 'freezer', 'walllamp', 'nail_polish', 'knifeblock', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'button', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'mousepad', 'pillow', 'wall', 'oven', 'powersocket', 'bench', 'cup', 'food_food', 'food_rice', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'table', 'board_game']\nTask:Grab some juice\nresponse:\n{\n    \"thoughts\": I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'toothbrush', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'mat', 'faucet', 'window', 'towel', 'pillow', 'wall', 'wall_clock', 'oven', 'bench', 'cutting_board', 'television', 'food_rice', 'ice', 'wooden_spoon', 'tray', 'comb', 'sink', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "Put toothpaste on toothbrush, brush teeth, rinse mouth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>", "[PUTOBJBACK] <tooth_paste>", "[WALK] <sink>", "[FIND] <teeth>", "[SCRUB] <teeth>", "[SCRUB] <teeth>", "[SCRUB] <teeth>", "[PUTBACK] <toothbrush> <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[RINSE] <teeth>", "[SWITCHOFF] <faucet>", "[GRAB] <toothbrush>", "[PUTOBJBACK] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'microwave', 'toilet', 'cat', 'wallshelf', 'bathroom_counter', 'floor', 'dvd_player', 'bookshelf', 'stovefan', 'ceilinglamp', 'toaster', 'mat', 'shower', 'faucet', 'tvstand', 'couch', 'towel', 'wall', 'bench', 'cutting_board', 'bed', 'razor', 'television', 'curtain', 'kitchen_counter', 'dresser', 'home_office', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Pet Cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'food_salt', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'powersocket', 'toy', 'tape', 'light_bulb', 'check', 'food_food', 'television', 'curtain', 'doorjamb', 'mouse', 'pencil', 'dresser', 'conditioner', 'spoon', 'light', 'table', 'colander']\nTask:Re arrange office\nresponse:\n", "output": {"thoughts": "Walk in.Turn on light.Move furniture. Change the room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[PUSH] <desk>", "[WALK] <chair>", "[PUSH] <chair>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'jelly', 'freezer', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'food_cake', 'mop_bucket', 'stovefan', 'mat', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'food_food', 'television', 'wine', 'tray', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'light', 'clothes_skirt']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Go to kitchen, open fridge, get groceries, put groceries in fridge, close fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'bathroom', 'food_carrot', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'stamp', 'faucet', 'stereo', 'shower', 'oil', 'closetdrawer', 'microphone', 'clothes_hat', 'wall', 'nightstand', 'oven', 'tape', 'cup', 'keys', 'toilet_paper', 'curtain', 'shoes', 'maindoor', 'sink', 'bathtub', 'food_steak', 'light', 'coffee']\nTask:Take dishes out of dishwasher\nresponse:\n", "output": {"thoughts": "Take clean dishes out of the dishwasher and put away", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'dog', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'form', 'wall', 'powersocket', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'mouse', 'cupboard', 'bathtub', 'sink', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": go to home office. turn the computer on. begin working,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_ice_cream', 'toothbrush', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'alcohol', 'brush', 'shoe-shine_kit', 'faucet', 'mousepad', 'closetdrawer', 'candle', 'clothes_pants', 'wall', 'standingmirror', 'food_food', 'television', 'tray', 'coffee_pot', 'towel_rack', 'dresser', 'bathtub', 'bathroom_cabinet', 'vase']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Using soft brush, carefully dust both Indonesian wedding doll statues.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <brush>", "[GRAB] <brush>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'filing_cabinet', 'tablelamp', 'crayon', 'woman', 'wallshelf', 'shampoo', 'floor', 'phone', 'desk', 'man', 'bookshelf', 'closetdrawer', 'couch', 'dining_room', 'nightstand', 'child', 'bed', 'photoframe', 'standingmirror', 'toilet_paper', 'oven_mitts', 'towel_rack', 'kitchen_counter', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'vase']\nTask:Say goodbye to guests leaving\nresponse:\n{\n    \"thoughts\": Tell them bye,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[FIND] <man>', '[TOUCH] <man>', '[FIND] <woman>', '[TOUCH] <woman>']\n}\n", "input": "\nScene_Objects:['food_bread', 'chair', 'cpuscreen', 'ceiling', 'character', 'video_game_console', 'bookshelf', 'keyboard', 'computer', 'mat', 'faucet', 'window', 'mousepad', 'wall_clock', 'tape', 'cutting_board', 'food_cheese', 'check', 'video_game_controller', 'mouthwash', 'food_onion', 'rag', 'coffee_pot', 'towel_rack', 'pencil', 'vacuum_cleaner', 'cd', 'dresser', 'home_office', 'dirt']\nTask:Playing video game\nresponse:\n", "output": {"thoughts": "Playing  video games  on various video game  systems such as wii u  and  xbox", "actions": ["[WALK] <home_office>", "[WALK] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[PUTOBJBACK] <video_game_controller>", "[SWITCHOFF] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'scissors', 'measuring_cup', 'freezer', 'food_butter', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'food_cake', 'mat', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'food_food', 'check', 'television', 'water', 'kitchen_counter', 'cupboard', 'door', 'light', 'clothes_skirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to kitchen, find water glass and grab it, find fridge and open grab water and pour into it, then drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[FIND] <freezer>', '[OPEN] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[CLOSE] <freezer>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'bathroom', 'milk', 'floor', 'bookshelf', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'bench', 'oven', 'toy', 'cutting_board', 'shelf', 'cup', 'standingmirror', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'console', 'pencil', 'comb', 'sink', 'home_office', 'scrabble']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Grab toys off floor.  Put it in a container.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'computer', 'button', 'tvstand', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'oven', 'coffee_table', 'novel', 'television', 'curtain', 'doorjamb', 'lighter', 'dresser', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Go to the living room, find the remote control on the coffee table, and change the TV channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <coffee_table>', '[WALK] <coffee_table>', '[TURNTO] <coffee_table>', '[LOOKAT] <coffee_table>', '[FIND] <remote_control>', '[TOUCH] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[TOUCH] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'bathroom', 'food_butter', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'shoe-shine_kit', 'shower', 'mousepad', 'closetdrawer', 'clothes_hat', 'wall', 'powersocket', 'cup', 'photoframe', 'novel', 'doorjamb', 'mouse', 'dresser', 'clothes_dress', 'light', 'clothes_skirt', 'vase']\nTask:Sew\nresponse:\n", "output": {"thoughts": "Using  the sewing machine  to  fix  and make  clothes", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[CLOSE] <dresser>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'ceilingfan', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'face', 'mail', 'table_cloth', 'faucet', 'food_jam', 'towel', 'dining_room', 'wall', 'face_soap', 'toy', 'photoframe', 'keys', 'water', 'rag', 'pencil', 'shoes', 'dresser', 'sink', 'bathroom_cabinet', 'home_office', 'colander', 'hands_both']\nTask:Wash face\nresponse:\n{\n    \"thoughts\": Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <rag> <sink>', '[GRAB] <rag>', '[FIND] <face_soap>', '[GRAB] <face_soap>', '[POUR] <face_soap> <rag>', '[PUTOBJBACK] <face_soap>', '[FIND] <face>', '[SCRUB] <face>', '[PUTBACK] <rag> <sink>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <hands_both>', '[POUR] <water> <hands_both>', '[RINSE] <face>', '[WASH] <rag>', '[GRAB] <rag>', '[PUTOBJBACK] <rag>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <face>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['mop', 'walllamp', 'food_butter', 'food_carrot', 'drawing', 'cpuscreen', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'food_bacon', 'electric_shaver', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'photoframe', 'curtain', 'rag', 'mouse', 'shoes', 'dresser', 'hairdryer', 'bag', 'light', 'tooth_paste']\nTask:Take off shoes\nresponse:\n", "output": {"thoughts": "Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <shoes> <dresser>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[PUTIN] <shoes> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'stamp', 'envelope', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'powersocket', 'standingmirror', 'television', 'curtain', 'doorjamb', 'mouse', 'bag', 'home_office', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to my living room and switch on the television, and lay on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'tablelamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'purse', 'computer', 'clothes_jacket', 'toaster', 'stereo', 'mousepad', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'bed', 'novel', 'mouse', 'shoes', 'dresser', 'clothes_dress', 'light', 'table']\nTask:Get ready for day\nresponse:\n", "output": {"thoughts": "Grab shoes, purse etc and head out the door!", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTON] <clothes_dress>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>", "[FIND] <purse>", "[GRAB] <purse>", "[CLOSE] <dresser>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['band-aids', 'trashcan', 'wallshelf', 'clothes_underwear', 'desk', 'keyboard', 'mail', 'computer', 'tvstand', 'shower', 'window', 'clothes_socks', 'couch', 'food_peanut_butter', 'pillow', 'basket_for_clothes', 'keys', 'coffe_maker', 'television', 'food_donut', 'tray', 'rag', 'headset', 'clothes_scarf', 'bathtub', 'sink', 'light', 'dishwasher', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['knife', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'folder', 'desk', 'keyboard', 'computer', 'mat', 'stereo', 'mousepad', 'closetdrawer', 'clothes_pants', 'wall', 'powersocket', 'cup', 'newspaper', 'novel', 'tray', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'spoon', 'light', 'soap', 'board_game', 'clothes_shirt', 'scrabble']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "Pick out the clothes you are going to wear and put them on your body", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'microwave', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'bench', 'cup', 'doorjamb', 'mouse', 'address_book', 'hairbrush', 'hairdryer', 'light', 'creditcard']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on computer. Open the address book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[FIND] <address_book>']\n}\n", "input": "\nScene_Objects:['juice', 'blender', 'beer', 'microwave', 'freezer', 'walllamp', 'detergent', 'floor', 'ceiling', 'character', 'stovefan', 'faucet', 'couch', 'clothes_hat', 'food_orange', 'broom', 'wall', 'oven', 'cutting_board', 'shelf', 'food_cheese', 'bed', 'coffe_maker', 'food_food', 'tray', 'after_shave', 'lighter', 'kitchen_counter', 'vase', 'coffee']\nTask:Eat cheese\nresponse:\n", "output": {"thoughts": "Open fridge.  Get cheese.  Eat it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[EAT] <food_cheese>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'chair', 'dog', 'freezer', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'pasta', 'light_bulb', 'clothes_gloves', 'television', 'curtain', 'doorjamb', 'conditioner', 'door', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room and turn on the light. I walk to the couch then sit on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['spectacles', 'microwave', 'chair', 'dog', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'laptop', 'mat', 'mousepad', 'closetdrawer', 'couch', 'wall', 'powersocket', 'coffe_maker', 'wine', 'doorjamb', 'kettle', 'mouse', 'pencil', 'dresser', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk in. Turn on the light. Set in my chair. Turn on lap top. And start to work.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <laptop>", "[GRAB] <laptop>", "[SWITCHON] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'mop', 'bathroom', 'dough', 'knifeblock', 'hanger', 'keyboard', 'alcohol', 'toaster', 'shower', 'foundation', 'couch', 'bowl', 'wall', 'pasta', 'basket_for_clothes', 'check', 'television', 'razor', 'tray', 'doorjamb', 'chessboard', 'dishrack', 'hairbrush', 'napkin', 'bathroom_cabinet', 'light', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": go to living room, find phone, pick up phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'freezer', 'nail_polish', 'woman', 'bathroom_counter', 'floor', 'shampoo', 'man', 'keyboard', 'purse', 'stovefan', 'mat', 'window', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'wall_clock', 'bench', 'shelf', 'standingmirror', 'check', 'television', 'curtain', 'doorjamb', 'pencil', 'bag', 'table', 'soap']\nTask:Hang with friends\nresponse:\n", "output": {"thoughts": "Sit with friends. Talk", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'milk', 'floor', 'ceiling', 'desk', 'keyboard', 'button', 'mop_bucket', 'toaster', 'stereo', 'window', 'wall', 'powersocket', 'bench', 'newspaper', 'coffe_maker', 'tray', 'curtain', 'food_sugar', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'coffee_cup', 'dishwasher', 'light', 'clothes_shirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[LOOKAT] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <milk>', '[GRAB] <milk>', '[PUTBACK] <milk> <coffe_maker>', '[FIND] <food_sugar>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[SWITCHON] <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'walllamp', 'woman', 'drawing', 'phone', 'desk', 'man', 'toaster', 'laptop', 'mat', 'shower', 'electric_shaver', 'oil', 'form', 'pillow', 'dining_room', 'oven', 'shelf', 'bed', 'standingmirror', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'tooth_paste', 'scrabble']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'nail_polish', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'bedroom', 'tea_bag', 'laptop', 'tvstand', 'clothes_socks', 'closetdrawer', 'couch', 'clothes_hat', 'wall', 'nightstand', 'powersocket', 'photoframe', 'coffe_maker', 'check', 'television', 'doorjamb', 'towel_rack', 'cards', 'dresser', 'door', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the dining room and turn on the lights,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'dough', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'window', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'check', 'curtain', 'mouse', 'pencil', 'dresser', 'light', 'table', 'vase', 'shoe_rack']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Put child in bed, sit down,read book.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'knifeblock', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'desk', 'food_cake', 'stovefan', 'toaster', 'deck_of_cards', 'faucet', 'candle', 'pillow', 'dining_room', 'oven', 'photoframe', 'check', 'food_food', 'television', 'bathroom_cabinet', 'table', 'dishwasher', 'soap', 'board_game', 'clothes_shirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find cat. Walk to cat. Pull cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[PULL] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'food_salt', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'character', 'shower', 'stereo', 'microphone', 'wall', 'toy', 'cutting_board', 'cup', 'novel', 'food_food', 'check', 'food_onion', 'coffee_pot', 'console', 'shaving_cream', 'towel_rack', 'shoes', 'cd', 'hairbrush', 'maindoor', 'soap', 'coffee']\nTask:Read\nresponse:\n", "output": {"thoughts": "Read a book in the living room", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[CLOSE] <filing_cabinet>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'toilet', 'chair', 'bathroom', 'drawing', 'longboard', 'floor', 'ceiling', 'food_cake', 'mat', 'ceilinglamp', 'shower', 'deck_of_cards', 'balanceball', 'shoe-shine_kit', 'tvstand', 'candle', 'food_kiwi', 'wall', 'nightstand', 'powersocket', 'bench', 'food_cheese', 'novel', 'coffe_maker', 'curtain', 'towel_rack', 'home_office', 'dishwasher']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to the bathroom and wash with the shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>', '[WASH] <shower>']\n}\n", "input": "\nScene_Objects:['chair', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'tvstand', 'shower', 'closetdrawer', 'wall', 'wall_clock', 'oven', 'cup', 'standingmirror', 'check', 'television', 'centerpiece', 'console', 'mouse', 'vacuum_cleaner', 'shoes', 'comb', 'dresser', 'hairbrush', 'sink', 'bag', 'napkin', 'bathroom_cabinet', 'vase']\nTask:vacuum carpet\nresponse:\n", "output": {"thoughts": "I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'computer', 'stovefan', 'faucet', 'window', 'mousepad', 'closetdrawer', 'couch', 'towel', 'wall', 'nightstand', 'curtain', 'towel_rack', 'mouse', 'coffee_pot', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom and locate the sink. I turn the faucet so that water is running, and rinse my hands briefly. Then I grab the soap and squeeze some onto my hands. I spend a few minutes rubbing my hands together, then I rinse them again. Finally, I turn the faucet off and wipe my hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SQUEEZE] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['food_bread', 'tablelamp', 'blender', 'tea', 'chair', 'freezer', 'drawing', 'orchid', 'hanger', 'floor', 'shampoo', 'stamp', 'bedroom', 'desk', 'computer', 'mat', 'stereo', 'mousepad', 'food_jam', 'dining_room', 'wall', 'wall_clock', 'bench', 'photoframe', 'food_food', 'dresser', 'napkin', 'home_office', 'table', 'spoon']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'deck_of_cards', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'door', 'light', 'table']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Keep my keyboard vacuumed and free of debris.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bathroom', 'detergent', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'thread', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'photoframe', 'toilet_paper', 'television', 'razor', 'coin', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Push power button.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'plate', 'couch', 'pillow', 'wall', 'newspaper', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'bathtub', 'light', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": My cat looks like it wants attention, so I will go pet it on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'band-aids', 'freezer', 'orchid', 'cpuscreen', 'floor', 'bookshelf', 'envelope', 'toaster', 'stovefan', 'plate', 'shower', 'faucet', 'foundation', 'oil', 'mousepad', 'food_jam', 'microphone', 'dining_room', 'bench', 'oven', 'toy', 'box', 'food_food', 'food_onion', 'diary', 'kitchen_counter', 'cupboard', 'colander']\nTask:Get out dish\nresponse:\n", "output": {"thoughts": "Open cupboard.  Get out the dish.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'soap']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to my home office, sit on the chair and turn on the computer. I open my email account to look for all the information I need.I pull up my online task on the computer to look at what the assigned task is.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'sheets', 'food_salt', 'tea', 'blow_dryer', 'wallshelf', 'ceiling', 'character', 'food_cake', 'computer', 'tvstand', 'window', 'mousepad', 'cutting_board', 'basket_for_clothes', 'cup', 'newspaper', 'light_bulb', 'novel', 'photoframe', 'coin', 'food_donut', 'rag', 'cards', 'kitchen_counter', 'vacuum_cleaner', 'pajamas', 'maindoor', 'bag', 'creditcard']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Clean the welcome mat", "actions": ["[WALK] <home_office>", "[WALK] <vacuum_cleaner>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'light_bulb', 'novel', 'television', 'curtain', 'cupboard', 'bathtub', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the living room, take a book from the shelf, sit in the sofa and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'fork', 'scissors', 'food_salt', 'dog', 'milk', 'drawing', 'needle', 'hanger', 'bedroom', 'keyboard', 'bookshelf', 'stovefan', 'tvstand', 'closetdrawer', 'wall_clock', 'oven', 'toy', 'photoframe', 'novel', 'toilet_paper', 'food_food', 'check', 'food_onion', 'kitchen_counter', 'hairdryer', 'bathtub', 'vase', 'instrument_violin', 'creditcard']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the book shelf, get a book, open the book and read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'powersocket', 'clothes_gloves', 'television', 'razor', 'curtain', 'doorjamb', 'console', 'mouse', 'door', 'bag', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk in living room. I find TV remote. I grab TV remote. I sit on sofa. I switch on TV. I switch channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'phone', 'button', 'mat', 'stereo', 'window', 'homework', 'foundation', 'couch', 'pillow', 'wall', 'toy', 'cup', 'light_bulb', 'television', 'razor', 'curtain', 'after_shave', 'doorjamb', 'rag', 'coffee_pot', 'laundry_detergent', 'bathtub', 'table']\nTask:Arrange furniture\nresponse:\n", "output": {"thoughts": "Move furniture where you want it.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[PUSH] <coffee_table>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'measuring_cup', 'toilet', 'chair', 'cat', 'milk', 'orchid', 'phone', 'video_game_console', 'bookshelf', 'keyboard', 'purse', 'table_cloth', 'brush', 'mat', 'iron', 'couch', 'cup', 'basket_for_clothes', 'video_game_controller', 'oven_mitts', 'food_donut', 'doorjamb', 'console', 'dresser', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Playing video game\nresponse:\n{\n    \"thoughts\": Playing  video games  on various video game  systems such as wii u  and  xbox,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <video_game_console>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[PUTOBJBACK] <video_game_controller>', '[SWITCHOFF] <video_game_console>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'ground_coffee', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'bookmark', 'ceilinglamp', 'window', 'foundation', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'powersocket', 'television', 'coffee_filter', 'doorjamb', 'mouse', 'dresser', 'maindoor', 'sink', 'light', 'table', 'shoe_rack']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Compose an email in the home office.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'bathroom', 'orchid', 'wallshelf', 'ceiling', 'phone', 'stamp', 'bookshelf', 'mail', 'computer', 'tvstand', 'shower', 'couch', 'pillow', 'oven', 'powersocket', 'box', 'bed', 'clothes_gloves', 'photoframe', 'check', 'curtain', 'pencil', 'dresser', 'cupboard', 'door', 'sink', 'bathtub', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find a cat, touch it gently,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'dog', 'band-aids', 'walllamp', 'ground_coffee', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'phone', 'food_cake', 'computer', 'toaster', 'mat', 'faucet', 'window', 'couch', 'candle', 'pillow', 'wall', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'table', 'light', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'food_salt', 'freezer', 'drawing', 'knife', 'cpuscreen', 'knifeblock', 'floor', 'character', 'bookshelf', 'laptop', 'plate', 'mousepad', 'candle', 'dining_room', 'wall', 'light_bulb', 'food_food', 'coffe_maker', 'tray', 'kitchen_counter', 'chessboard', 'dishrack', 'shoes', 'cupboard', 'laundry_detergent', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'food_butter', 'dough', 'food_carrot', 'drawing', 'food_egg', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'food_chicken', 'food_cake', 'stovefan', 'mat', 'faucet', 'oil', 'wall', 'pasta', 'powersocket', 'oven', 'wall_clock', 'cutting_board', 'food_food', 'coffe_maker', 'oven_mitts', 'wooden_spoon', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'maindoor', 'sink', 'bathroom_cabinet', 'light']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and walk over the cabinent. I pull out a pan and walk to the stove. I put the pan on the stove and walk over to the fridge. I pull out out meat from the fridge and walk back over the stove. Then I put the meat in the pan.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[WALK] <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'ceilingfan', 'orchid', 'bathroom_counter', 'bookshelf', 'purse', 'stovefan', 'mat', 'iron', 'electric_shaver', 'food_noodles', 'mousepad', 'couch', 'clothes_pants', 'pillow', 'powersocket', 'keys', 'coffe_maker', 'oven_mitts', 'wine', 'shaving_cream', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'home_office', 'light', 'tooth_paste']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['microwave', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'mat', 'window', 'clothes_socks', 'couch', 'pillow', 'towel', 'broom', 'wall', 'bench', 'oven', 'toy', 'bed', 'coffe_maker', 'television', 'wooden_spoon', 'curtain', 'rag', 'doorjamb', 'food_onion', 'towel_rack', 'mouse', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Find couch. Sit on couch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'measuring_cup', 'microwave', 'walllamp', 'trashcan', 'floor', 'bedroom', 'computer', 'toaster', 'stovefan', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'candle', 'pillow', 'wall', 'nightstand', 'bench', 'oven', 'cup', 'newspaper', 'coffe_maker', 'tray', 'kitchen_counter', 'cupboard', 'home_office', 'dishwasher', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the sink. Take dishes out of sink. Open dishwasher and place dishes in dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'floor', 'dvd_player', 'bedroom', 'bookshelf', 'food_cake', 'shower', 'picture', 'foundation', 'mousepad', 'closetdrawer', 'tooth_paste', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'shelf', 'food_food', 'doorjamb', 'coffee_pot', 'mouse', 'pencil', 'kitchen_counter', 'towel_rack', 'door', 'napkin', 'bathroom_cabinet', 'board_game']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I go to the bathroom. I locate the toilet. I walk to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[WALK] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'walllamp', 'ceilingfan', 'drawing', 'hanger', 'floor', 'bedroom', 'computer', 'ceilinglamp', 'deck_of_cards', 'iron', 'window', 'electric_shaver', 'couch', 'food_orange', 'wall', 'wall_clock', 'bed', 'novel', 'check', 'television', 'toothbrush_holder', 'wooden_spoon', 'after_shave', 'towel_rack', 'dresser', 'conditioner', 'maindoor', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'food_egg', 'bathroom_counter', 'ceiling', 'computer', 'button', 'food_bacon', 'plate', 'stereo', 'shower', 'window', 'picture', 'food_kiwi', 'wall', 'wall_clock', 'toy', 'shelf', 'bed', 'toilet_paper', 'oven_mitts', 'tray', 'after_shave', 'doorjamb', 'mouse', 'pencil', 'clothes_dress', 'maindoor', 'sink', 'food_steak', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into the living room and turn on the radio.  Then, I select the station I want, which is usually a preset.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'bathroom', 'ceilingfan', 'drawing', 'bills', 'knife', 'orchid', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'bookmark', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'couch', 'pillow', 'wall', 'oven', 'keys', 'television', 'coffee_filter', 'curtain', 'centerpiece', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": change the tv channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'tea', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'table_cloth', 'mat', 'stereo', 'window', 'foundation', 'mousepad', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'bench', 'keys', 'toilet_paper', 'food_food', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'vacuum_cleaner', 'comb', 'maindoor', 'bathtub', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'juice', 'toilet', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'phone', 'desk', 'stovefan', 'plate', 'faucet', 'window', 'mousepad', 'form', 'pillow', 'wall', 'wall_clock', 'oven', 'clothes_gloves', 'tray', 'diary', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'sink']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[TURNTO] <plate>', '[LOOKAT] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WASH] <plate>', '[PUTBACK] <plate> <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'walllamp', 'food_carrot', 'drawing', 'wallshelf', 'hanger', 'phone', 'character', 'desk', 'table_cloth', 'mat', 'tvstand', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall_clock', 'powersocket', 'cutting_board', 'television', 'food_donut', 'food_rice', 'towel_rack', 'controller', 'maindoor']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find a cat, touch it gently", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'toilet', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'toaster', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'love_seat', 'clothes_gloves', 'television', 'tray', 'curtain', 'food_onion', 'dresser', 'cupboard', 'instrument_guitar', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I'll watch TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'blender', 'milk', 'orchid', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'thread', 'stovefan', 'mat', 'toaster', 'faucet', 'plate', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'wall_clock', 'oven', 'tray', 'curtain', 'mouse', 'sink', 'napkin', 'bathroom_cabinet', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom, find sink, turn on faucet, put soap in hands, wash and rinse hands,", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <soap>", "[WASH] <soap>", "[RINSE] <soap>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'food_bread', 'dog', 'ground_coffee', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'bedroom', 'alcohol', 'mop_bucket', 'stovefan', 'laptop', 'tvstand', 'ceilinglamp', 'shoe-shine_kit', 'window', 'mat', 'mousepad', 'food_orange', 'pillow', 'powersocket', 'coffe_maker', 'fly', 'coffee_pot', 'kitchen_counter', 'cupboard', 'dishwasher', 'soap']\nTask:Watch fly\nresponse:\n{\n    \"thoughts\": Look around for a fly, often on a wall or ceiling.  Follow it's movements.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <fly>', '[FIND] <fly>', '[TURNTO] <fly>', '[LOOKAT] <fly>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'walllamp', 'orchid', 'hanger', 'phone', 'computer', 'piano_bench', 'nightstand', 'pasta', 'wall_clock', 'powersocket', 'toy', 'food_cheese', 'clothes_gloves', 'check', 'toilet_paper', 'television', 'after_shave', 'doorjamb', 'door', 'hairdryer', 'sink', 'home_office', 'light', 'table', 'clothes_scarf', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to bedroom, sit down, pick up book, read", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'stovefan', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'towel', 'wall', 'powersocket', 'photoframe', 'coffee_filter', 'coin', 'tray', 'food_onion', 'doorjamb', 'towel_rack', 'dishrack', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom. Turn on the sink. Rinse hands. Apply soap and scrub your hands together. Rinse hands again. Turn off faucet. Dry hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'thread', 'computer', 'piano_bench', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to the home office and turn to the desk. Switch on the computer. Grab the mouse and keyboard to start working.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[TURNTO] <desk>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'ceilinglamp', 'laptop', 'faucet', 'tvstand', 'window', 'couch', 'microphone', 'towel', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'novel', 'food_food', 'television', 'tray', 'mouse', 'dresser', 'sink', 'clothes_scarf', 'home_office', 'table', 'clothes_skirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'dog', 'walllamp', 'ground_coffee', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'shower', 'faucet', 'clothes_socks', 'towel', 'wall', 'cutting_board', 'shelf', 'basket_for_clothes', 'coffe_maker', 'food_rice', 'water', 'tray', 'doorjamb', 'food_steak', 'home_office', 'vase', 'soap', 'teeth']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[TURNTO] <shower>", "[FIND] <water>", "[RINSE] <water>", "[FIND] <soap>", "[WASH] <soap>", "[RINSE] <water>", "[FIND] <towel>", "[GRAB] <towel>", "[SCRUB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'dog', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'longboard', 'floor', 'phone', 'character', 'purse', 'computer', 'toaster', 'plate', 'balanceball', 'stereo', 'dish_soap', 'food_kiwi', 'wall', 'oven', 'tape', 'coffe_maker', 'mouthwash', 'tray', 'door', 'sink', 'napkin', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'oil', 'pillow', 'dining_room', 'wall', 'wall_clock', 'bench', 'standingmirror', 'television', 'curtain', 'doorjamb', 'mouse', 'headset', 'napkin', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <chair>", "[SIT] <chair>", "[TURNTO] <television>", "[WATCH] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'freezer', 'trashcan', 'toothbrush', 'orchid', 'hanger', 'desk', 'keyboard', 'bookmark', 'shoe-shine_kit', 'iron', 'tvstand', 'plate', 'mousepad', 'microphone', 'wall', 'bench', 'toilet_paper', 'television', 'toothbrush_holder', 'ice', 'shoes', 'chessboard', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'tooth_paste', 'teeth', 'creditcard']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the toothbrush holder, grab the toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now wash the teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'dough', 'cpuscreen', 'hanger', 'computer', 'piano_bench', 'bookmark', 'mat', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'wall', 'nightstand', 'box', 'photoframe', 'toilet_paper', 'video_game_controller', 'food_food', 'doorjamb', 'towel_rack', 'hairbrush', 'hair', 'instrument_violin', 'sink', 'home_office', 'light', 'vase', 'tooth_paste']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into room, grab phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'floor', 'ceiling', 'bedroom', 'window', 'clothes_socks', 'couch', 'dustpan', 'clothes_pants', 'food_orange', 'food_kiwi', 'wall', 'toy', 'bed', 'photoframe', 'food_food', 'coffe_maker', 'wine', 'doorjamb', 'diary', 'console', 'mouse', 'door', 'bathroom_cabinet', 'light', 'vase', 'soap']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Grasp light switch, flip light switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[TURNTO] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'ground_coffee', 'walllamp', 'dough', 'food_carrot', 'floor', 'ceiling', 'stovefan', 'toaster', 'stereo', 'clothes_hat', 'wall', 'oven', 'bench', 'cutting_board', 'food_cheese', 'coffe_maker', 'food_food', 'check', 'toothbrush_holder', 'food_donut', 'tray', 'kitchen_counter', 'dishrack', 'hairbrush', 'bag', 'bathroom_cabinet', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, put the groceries in the fridge and close it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'nail_polish', 'bathroom', 'detergent', 'knife', 'cpuscreen', 'dvd_player', 'bedroom', 'computer', 'table_cloth', 'mop_bucket', 'brush', 'toaster', 'shower', 'window', 'couch', 'dining_room', 'wall', 'pasta', 'cup', 'photoframe', 'after_shave', 'pencil', 'kitchen_counter', 'dresser', 'clothes_dress', 'sink', 'instrument_guitar', 'light', 'creditcard']\nTask:Sit\nresponse:\n{\n    \"thoughts\": We sit at the dining room table,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['glue', 'dog', 'bathroom', 'toothbrush', 'orchid', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'dining_room', 'wall', 'wall_clock', 'oven', 'razor', 'toothbrush_holder', 'tray', 'centerpiece', 'doorjamb', 'instrument_violin', 'sink', 'bathroom_cabinet', 'vase', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "go to bathroom. grab toothbrush. put toothpaste on toothbrush. brush teeth. turn on faucet. rinse toothbrush and mouth. put toothbrush away.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'freezer', 'trashcan', 'drawing', 'floor', 'phone', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'toaster', 'mat', 'faucet', 'window', 'wall', 'wall_clock', 'powersocket', 'box', 'cup', 'bed', 'video_game_controller', 'television', 'doorjamb', 'console', 'kitchen_counter', 'door', 'light', 'teeth']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk over to the phone and pick it up.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['spectacles', 'blender', 'crayon', 'microwave', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'bedroom', 'envelope', 'stovefan', 'window', 'oil', 'wall', 'powersocket', 'bench', 'toy', 'keys', 'food_donut', 'wooden_spoon', 'curtain', 'doorjamb', 'maindoor', 'bathtub', 'bathroom_cabinet', 'light', 'dirt', 'teeth']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_carrot', 'milk', 'dough', 'knifeblock', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'mat', 'faucet', 'shower', 'stereo', 'towel', 'pillow', 'wall', 'bench', 'cup', 'bed', 'photoframe', 'coffe_maker', 'wooden_spoon', 'after_shave', 'rag', 'mouse', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Turn on the water faucet. Wet my hands and turn off the faucet. Put soap on my hands. Rub hands together, making sure the soap covers all areas of the hands. Scrub hands for about 20 seconds. Turn on the water water faucet. Rinse hands under the running water. Turn off the water faucet. Dry hands with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'table_cloth', 'stovefan', 'mat', 'faucet', 'foundation', 'mousepad', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'toy', 'hands_both', 'tray', 'doorjamb', 'towel_rack', 'shaving_cream', 'door', 'sink', 'spoon', 'bathroom_cabinet', 'light', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I need to rinse my dentures off and put them back in my mouth.", "actions": ["[WALK] <bathroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <teeth>", "[GRAB] <teeth>", "[WASH] <teeth>", "[PUTOBJBACK] <teeth>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <towel>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'fork', 'tablelamp', 'walllamp', 'cpuscreen', 'bathroom_counter', 'floor', 'toaster', 'plate', 'tvstand', 'dish_soap', 'food_noodles', 'bowl', 'wall', 'pasta', 'oven', 'box', 'coffe_maker', 'television', 'ice', 'tray', 'curtain', 'console', 'towel_rack', 'pencil', 'maindoor', 'sink', 'instrument_guitar', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Wash a plate using the dishwasher in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'walllamp', 'nail_polish', 'cpuscreen', 'bathroom_counter', 'character', 'desk', 'folder', 'keyboard', 'food_cake', 'bedroom', 'alcohol', 'stovefan', 'toaster', 'shower', 'window', 'candle', 'pillow', 'wall', 'wall_clock', 'oven', 'nightstand', 'cutting_board', 'video_game_controller', 'curtain', 'doorjamb', 'towel_rack', 'shoes', 'board_game']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up your book and read it", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'freezer', 'walllamp', 'food_carrot', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'toaster', 'shower', 'wall', 'novel', 'food_food', 'television', 'oven_mitts', 'tray', 'kitchen_counter', 'headset', 'dresser', 'cupboard', 'hair', 'bathtub']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Go to kitchen, open fridge, get groceries, put groceries in fridge, close fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'toilet', 'detergent', 'toothbrush', 'orchid', 'wallshelf', 'hanger', 'bedroom', 'bookshelf', 'computer', 'shower', 'window', 'cutting_board', 'food_cheese', 'novel', 'check', 'mouthwash', 'toothbrush_holder', 'wine', 'doorjamb', 'pajamas', 'hairdryer', 'maindoor', 'bathtub', 'home_office', 'door', 'clothes_skirt', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, grab the toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now wash the teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'faucet', 'shower', 'tooth_paste', 'towel', 'wall', 'wall_clock', 'oven', 'tape', 'photoframe', 'clothes_gloves', 'light_bulb', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'light', 'vase', 'soap', 'home_office', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, open the faucet rinse the hands, wash the hands with soap and rinse again, then wipe with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['freezer', 'bathroom', 'food_carrot', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stovefan', 'mat', 'faucet', 'dining_room', 'wall', 'wall_clock', 'oven', 'newspaper', 'light_bulb', 'tray', 'curtain', 'towel_rack', 'laundry_detergent', 'door', 'sink', 'napkin', 'bathroom_cabinet', 'bathtub', 'home_office', 'soap', 'laser_pointer', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "go to the bathroom and wash your hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[SQUEEZE] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'nightstand', 'feet_both', 'photoframe', 'bed', 'television', 'coin', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'shoes', 'hairbrush', 'door']\nTask:Put on your shoes\nresponse:\n{\n    \"thoughts\": Sit on a chair with your shoes in front of you.  Put them on one at a time.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <feet_both>', '[PUTBACK] <shoes> <feet_both>']\n}\n", "input": "\nScene_Objects:['scissors', 'blender', 'tablelamp', 'ground_coffee', 'bathroom', 'toothbrush', 'drawing', 'knife', 'orchid', 'desk', 'keyboard', 'brush', 'shower', 'homework', 'food_jam', 'pillow', 'dining_room', 'bowl', 'wall_clock', 'check', 'television', 'toothbrush_holder', 'tray', 'kettle', 'kitchen_counter', 'maindoor', 'sink', 'clothes_skirt', 'tooth_paste', 'dirt']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'tablelamp', 'mop', 'microwave', 'freezer', 'walllamp', 'needle', 'food_egg', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'stovefan', 'stereo', 'form', 'dining_room', 'wall', 'oven', 'powersocket', 'cutting_board', 'toy', 'keys', 'food_food', 'coffe_maker', 'wine', 'tray', 'kitchen_counter', 'clothes_scarf', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Open the fridge, and put groceries in the fridge. Close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_bread', 'microwave', 'freezer', 'ground_coffee', 'dough', 'orchid', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'tea_bag', 'toaster', 'faucet', 'shower', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'curtain', 'kitchen_counter', 'cupboard', 'sink']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'microphone', 'form', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'coffe_maker', 'television', 'curtain', 'vacuum_cleaner', 'dishrack', 'dresser', 'hairbrush', 'door', 'table', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['scissors', 'crayon', 'chair', 'nail_polish', 'cat', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'bookshelf', 'computer', 'food_bacon', 'clothes_jacket', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'toy', 'doorjamb', 'kettle', 'mouse', 'dresser', 'hairdryer', 'sink', 'bathroom_cabinet', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Coming home turning on lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'microwave', 'freezer', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'button', 'toaster', 'stovefan', 'mat', 'faucet', 'shower', 'couch', 'wall', 'wall_clock', 'oven', 'keys', 'coffe_maker', 'water', 'tray', 'curtain', 'coffee_pot', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'sponge', 'microwave', 'freezer', 'walllamp', 'food_carrot', 'hanger', 'bathroom_counter', 'floor', 'envelope', 'stovefan', 'mat', 'faucet', 'clothes_hat', 'wall', 'wall_clock', 'oven', 'cutting_board', 'cup', 'food_vegetable', 'food_food', 'coin', 'water', 'tray', 'kitchen_counter', 'shoes', 'door', 'sink', 'bathroom_cabinet']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "walk to kicthen , find stove , find cooking pot , find vegetables , find water  ,put vegetables in cooking pot, pour water to cooking pot,put cooking pot on stove, switch on stove", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[FIND] <pot>", "[FIND] <food_vegetable>", "[FIND] <water>", "[GRAB] <water>", "[WASH] <food_vegetable>", "[POUR] <water> <pot>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'toilet', 'chair', 'walllamp', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'shower', 'mousepad', 'food_peanut_butter', 'food_jam', 'wall', 'bench', 'powersocket', 'toy', 'box', 'standingmirror', 'coffe_maker', 'water', 'doorjamb', 'cards', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'light', 'soap', 'board_game']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'knife', 'cpuscreen', 'hanger', 'floor', 'phone', 'desk', 'keyboard', 'thread', 'computer', 'piano_bench', 'toaster', 'laptop', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'box', 'console', 'mouse', 'dresser', 'maindoor', 'hairdryer', 'board_game']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I open my laptop and then login to Windows. Then I open Google Chrome, and type gmail.com into the address bar. My email automatically logs in, then I click the Compose button, type in the address of the recipient, then the subject of the email, and then start typing my email", "actions": ["[WALK] <laptop>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'detergent', 'milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'napkin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Find the remote, turn on the TV, watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'microwave', 'dog', 'drawing', 'hanger', 'bedroom', 'desk', 'computer', 'table_cloth', 'stovefan', 'brush', 'faucet', 'mousepad', 'dining_room', 'bowl', 'wall_clock', 'powersocket', 'novel', 'food_food', 'check', 'coffee_filter', 'kitchen_counter', 'cd', 'food_steak', 'napkin', 'home_office', 'light', 'clothes_skirt', 'vase', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the radio. Turn on the radio. Listen to the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'kettle', 'mouse', 'dresser', 'hair', 'door', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": walk to living room. look at computer. switch on computer. sit in chair. watch computer. switch off computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[WATCH] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'walllamp', 'drawing', 'cpuscreen', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'alcohol', 'laptop', 'plate', 'faucet', 'homework', 'mousepad', 'dining_room', 'check', 'food_food', 'television', 'tray', 'wooden_spoon', 'mouse', 'shoes', 'conditioner', 'maindoor', 'napkin', 'home_office', 'soap', 'instrument_violin']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I turn on computer. I use keyboard to type. I open browser. I search.", "actions": ["[WALK] <bedroom>", "[WALK] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'sheets', 'mop', 'microwave', 'chair', 'freezer', 'blow_dryer', 'nail_polish', 'orchid', 'knife', 'floor', 'keyboard', 'stovefan', 'stereo', 'foundation', 'couch', 'soap', 'food_orange', 'pillow', 'wall', 'powersocket', 'check', 'toilet_paper', 'coffe_maker', 'curtain', 'pencil', 'shoes', 'vacuum_cleaner', 'colander', 'shoe_rack']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Petting the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'jelly', 'freezer', 'chair', 'blow_dryer', 'bathroom', 'walllamp', 'knife', 'phone', 'keyboard', 'stovefan', 'homework', 'couch', 'food_jam', 'bench', 'oven', 'toy', 'box', 'food_cheese', 'oven_mitts', 'food_rice', 'tray', 'curtain', 'after_shave', 'dresser', 'clothes_dress', 'conditioner', 'hairdryer', 'vase', 'board_game']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my living room, find my cat, scrub it using my hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'food_ice_cream', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'toaster', 'stovefan', 'mat', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'wall', 'oven', 'powersocket', 'cutting_board', 'coffe_maker', 'television', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'hair', 'door', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I enter the kitchen and grab a waterglass from the cupboard. I open the fridge and grap some water. I pour the water into the cup and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[OPEN] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['toilet', 'cat', 'floor', 'character', 'stamp', 'table_cloth', 'piano_bench', 'mat', 'window', 'couch', 'food_jam', 'pillow', 'wall', 'nightstand', 'bench', 'toy', 'cup', 'mouthwash', 'television', 'coffee_filter', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'clothes_dress', 'door', 'food_steak', 'bathtub', 'bag', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I see my cat on the couch so I walk over, sit down and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom', 'ceilingfan', 'food_carrot', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'faucet', 'shower', 'window', 'wall', 'photoframe', 'novel', 'toilet_paper', 'curtain', 'doorjamb', 'diary', 'towel_rack', 'clothes_dress', 'sink', 'bathroom_cabinet', 'instrument_guitar', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk in to the bathroom. I find the soap. I find the faucet. I turn on the faucet. I wash my hands. I put the soap back. I rinse my hands. I turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['fork', 'blender', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'keyboard', 'mat', 'laptop', 'tvstand', 'picture', 'window', 'clothes_socks', 'couch', 'pillow', 'food_kiwi', 'wall', 'wall_clock', 'cutting_board', 'bed', 'television', 'toothbrush_holder', 'curtain', 'food_onion', 'doorjamb', 'door', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I cross the room to the couch. I sit down and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'tea_bag', 'mat', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'toy', 'photoframe', 'check', 'mouthwash', 'food_food', 'food_donut', 'wine', 'curtain', 'doorjamb', 'console', 'dresser', 'door', 'light']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'bathroom', 'detergent', 'food_butter', 'drawing', 'wallshelf', 'floor', 'desk', 'bookshelf', 'plate', 'faucet', 'picture', 'iron', 'window', 'dish_soap', 'food_jam', 'wall', 'nightstand', 'toy', 'cutting_board', 'food_food', 'television', 'curtain', 'pajamas', 'conditioner', 'light', 'dishwasher', 'tooth_paste', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'mop', 'toilet', 'chair', 'trashcan', 'knifeblock', 'hanger', 'longboard', 'floor', 'phone', 'keyboard', 'balanceball', 'foundation', 'pillow', 'wall', 'powersocket', 'box', 'photoframe', 'bed', 'check', 'rag', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'maindoor', 'light', 'colander']\nTask:Sit\nresponse:\n{\n    \"thoughts\": We sit at the dining room table,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'orchid', 'hanger', 'ceiling', 'keyboard', 'piano_bench', 'tvstand', 'shower', 'electric_shaver', 'clothes_socks', 'couch', 'wall', 'powersocket', 'bench', 'cutting_board', 'toy', 'photoframe', 'bed', 'razor', 'wooden_spoon', 'after_shave', 'coffee_pot', 'dresser', 'clothes_dress', 'bathtub', 'bathroom_cabinet', 'light', 'clothes_shirt', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "walk to kitchen. grab clothes_dress. look at washing machine. open washing machine. put clothes_dress on washing machine. close washing machine. switch on washing machine. switch off washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <clothes_dress>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[TURNTO] <washing_machine>", "[LOOKAT] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>", "[SWITCHOFF] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'fork', 'beer', 'walllamp', 'bathroom', 'orchid', 'kitchen_cabinet', 'floor', 'phone', 'toaster', 'mat', 'plate', 'tvstand', 'window', 'dish_soap', 'couch', 'bowl', 'wall', 'oven', 'bench', 'coffe_maker', 'tray', 'rag', 'console', 'kitchen_counter', 'dishrack', 'hair', 'maindoor', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>', '[GRAB] <plate>', '[WIPE] <plate>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[PUTOBJBACK] <dishrack>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'detergent', 'cpuscreen', 'ceiling', 'character', 'thread', 'food_bacon', 'mat', 'shower', 'window', 'candle', 'pillow', 'dining_room', 'oven', 'powersocket', 'tape', 'cutting_board', 'clothes_gloves', 'coffe_maker', 'check', 'television', 'food_onion', 'cards', 'towel_rack', 'pencil', 'maindoor', 'home_office', 'table', 'teeth']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "go to living room, find phone, pick up phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'microwave', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'button', 'food_bacon', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'coffee_table', 'novel', 'mouthwash', 'television', 'doorjamb', 'diary', 'clothes_scarf', 'sink']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit down in recliner.  Pick up novel off of coffee table.  Open novel to last read page.  Begin reading,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <bookmark>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'shoe_rack', 'filing_cabinet', 'beer', 'microwave', 'freezer', 'nail_polish', 'drawing', 'stamp', 'keyboard', 'computer', 'bookmark', 'plate', 'tvstand', 'dish_soap', 'form', 'wall_clock', 'bench', 'toy', 'newspaper', 'novel', 'coffe_maker', 'food_donut', 'curtain', 'kitchen_counter', 'headset', 'cd', 'door', 'dishwasher', 'creditcard']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'folder', 'bookshelf', 'remote_control', 'button', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'sink', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go the living room, look for my tv remote control, take the control and then sit down on the couch and turn on the TV and push the buttons on the remote control to change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'beer', 'microwave', 'freezer', 'walllamp', 'blow_dryer', 'detergent', 'cat', 'milk', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'envelope', 'stovefan', 'couch', 'food_orange', 'food_kiwi', 'wall', 'oven', 'cutting_board', 'toy', 'shelf', 'coffe_maker', 'food_food', 'tray', 'kitchen_counter', 'maindoor', 'vase']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk to the kitchen, open the fridge and put the groceries in the fridge. Then I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'walllamp', 'ceilingfan', 'food_carrot', 'toothbrush', 'drawing', 'desk', 'bookshelf', 'stovefan', 'balanceball', 'shower', 'window', 'candle', 'clothes_pants', 'food_orange', 'dining_room', 'bench', 'bed', 'keys', 'food_food', 'diary', 'mouse', 'shoes', 'sink', 'vase', 'tooth_paste', 'scrabble', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": EARLY MORNING AND BEFORE GOING TO BED,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['juice', 'nail_polish', 'wallshelf', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'piano_bench', 'stovefan', 'mat', 'plate', 'faucet', 'picture', 'tvstand', 'dish_soap', 'food_noodles', 'pillow', 'food_kiwi', 'wall', 'wall_clock', 'oven', 'tape', 'shelf', 'coin', 'tray', 'curtain', 'sink', 'napkin', 'bathroom_cabinet']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Make some dish water, grab a dish rag, put dishes in dish water and wash.", "actions": ["[RUN] <sink>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[PUTBACK] <dish_soap> <plate>", "[RINSE] <dish_soap>", "[FIND] <plate>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ground_coffee', 'nail_polish', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'button', 'mail', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'tape', 'television', 'curtain', 'food_onion', 'cupboard', 'table', 'shoe_rack']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to the living room, sit on the couch, find the remote, switch on the TV and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'shampoo', 'bedroom', 'bookshelf', 'thread', 'mop_bucket', 'tvstand', 'faucet', 'mousepad', 'towel', 'powersocket', 'bed', 'coffe_maker', 'television', 'oven_mitts', 'tray', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'maindoor', 'home_office', 'bathroom_cabinet', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter living room. Walk to telephone. Pick up from hook.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'band-aids', 'walllamp', 'nail_polish', 'drawing', 'bills', 'wallshelf', 'orchid', 'hanger', 'knife', 'floor', 'ceiling', 'desk', 'alcohol', 'mat', 'foundation', 'closetdrawer', 'couch', 'pillow', 'wall', 'bed', 'novel', 'check', 'oven_mitts', 'rag', 'kitchen_counter', 'dresser', 'bag', 'napkin']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the book case, I select a book, I go back and sit down, then I read.,\n    \"actions\": ['[WALK] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'freezer', 'walllamp', 'drawing', 'hanger', 'floor', 'mat', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'cutting_board', 'cup', 'novel', 'food_food', 'television', 'wine', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'shoes', 'bathtub', 'bathroom_cabinet', 'table', 'soap', 'board_game']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "lie in the sofa, relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'toilet', 'freezer', 'walllamp', 'drawing', 'character', 'bedroom', 'toaster', 'deck_of_cards', 'faucet', 'shower', 'plate', 'window', 'mousepad', 'couch', 'candle', 'clothes_hat', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'tray', 'mouse', 'pencil', 'bathtub', 'bathroom_cabinet', 'table', 'dishwasher', 'board_game']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into the bedroom. I walk to stereo. I turn on stereo.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'chair', 'food_carrot', 'hanger', 'bathroom_counter', 'floor', 'keyboard', 'mat', 'plate', 'ceilinglamp', 'window', 'mousepad', 'couch', 'food_orange', 'pillow', 'broom', 'wall', 'toy', 'bed', 'toilet_paper', 'television', 'wine', 'curtain', 'food_onion', 'doorjamb', 'dresser', 'spoon', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Sit on sofa. Lie down on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'food_salt', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'desk', 'alcohol', 'stovefan', 'mat', 'plate', 'shower', 'couch', 'food_orange', 'pillow', 'wall', 'oven', 'food_food', 'television', 'tray', 'towel_rack', 'kitchen_counter', 'sauce_pan', 'cupboard', 'spoon', 'napkin', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <spoon>', '[GRAB] <spoon>', '[FIND] <sauce_pan>', '[PUTBACK] <spoon> <sauce_pan>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <plate>', '[PUTBACK] <pot> <plate>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'juice', 'food_salt', 'microwave', 'ground_coffee', 'nail_polish', 'drawing', 'food_egg', 'bathroom_counter', 'shampoo', 'purse', 'stovefan', 'shower', 'stereo', 'window', 'mousepad', 'towel', 'nightstand', 'oven', 'powersocket', 'curtain', 'after_shave', 'doorjamb', 'coffee_pot', 'mouse', 'dresser', 'hair', 'door', 'bathtub', 'sink']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room , find phone ,pick up phone 3", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'man', 'bookshelf', 'clothes_jacket', 'piano_bench', 'toaster', 'stereo', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'coffe_maker', 'doorjamb', 'cupboard', 'door', 'sink', 'home_office', 'table', 'light']\nTask:Have conversation with boyfriend\nresponse:\n{\n    \"thoughts\": Sit on sofa, talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'broom', 'wall', 'pasta', 'powersocket', 'shelf', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'light', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the computer and turn it on. I look at the computer screen. I move and click the mouse to select buttons on the screen.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[PUSH] <mouse>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'tablelamp', 'walllamp', 'ground_coffee', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'mat', 'faucet', 'couch', 'pillow', 'wall', 'pasta', 'wall_clock', 'cup', 'photoframe', 'bed', 'television', 'water', 'wine', 'curtain', 'paper_towel', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. I turn on the water. I put soap on my hands. I rinse my hands. I turn off the water. I dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <water>', '[TURNTO] <water>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <water>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>']\n}\n", "input": "\nScene_Objects:['mop', 'microwave', 'bathroom', 'detergent', 'wallshelf', 'clothes_underwear', 'hanger', 'ceiling', 'bookshelf', 'tvstand', 'faucet', 'deck_of_cards', 'window', 'clothes_socks', 'clothes_pants', 'cup', 'standingmirror', 'photoframe', 'check', 'tray', 'kitchen_counter', 'hair', 'laundry_detergent', 'door', 'bathroom_cabinet', 'tooth_paste', 'clothes_shirt', 'washing_machine', 'coffee', 'shoe_rack']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[DROP] <basket_for_clothes>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[GRAB] <laundry_detergent>", "[PUTOBJBACK] <laundry_detergent>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'detergent', 'orchid', 'wallshelf', 'floor', 'ceiling', 'tea_bag', 'stovefan', 'toaster', 'faucet', 'shower', 'foundation', 'tooth_paste', 'closetdrawer', 'candle', 'nightstand', 'cup', 'coffe_maker', 'food_food', 'television', 'coffee_filter', 'tray', 'curtain', 'rag', 'clothes_scarf', 'table', 'soap', 'instrument_violin', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'tablelamp', 'food_ice_cream', 'toilet', 'bathroom', 'toothbrush', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceilinglamp', 'faucet', 'clothes_socks', 'closetdrawer', 'tape', 'cup', 'check', 'food_food', 'razor', 'food_onion', 'centerpiece', 'mouse', 'dresser', 'bathtub', 'bathroom_cabinet', 'light', 'clothes_skirt', 'tooth_paste', 'shoe_rack']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "EARLY MORNING AND BEFORE GOING TO BED", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'microwave', 'ground_coffee', 'knifeblock', 'wallshelf', 'ceiling', 'phone', 'character', 'desk', 'piano_bench', 'stovefan', 'toaster', 'couch', 'clothes_hat', 'pillow', 'bench', 'oven', 'tape', 'cutting_board', 'doorjamb', 'cards', 'mouse', 'dishrack', 'cupboard', 'hairdryer', 'sink', 'clothes_scarf', 'table', 'dishwasher']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": i go find the phone. i pick up the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'sponge', 'beer', 'chair', 'freezer', 'detergent', 'food_butter', 'cpuscreen', 'hanger', 'bedroom', 'desk', 'bookshelf', 'table_cloth', 'toaster', 'ceilinglamp', 'shower', 'iron', 'window', 'wall', 'oven', 'food_food', 'razor', 'food_rice', 'curtain', 'kitchen_counter', 'food_steak', 'door', 'spoon', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'band-aids', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'iron', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'television', 'doorjamb', 'coffee_pot', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Go to home office, turn on the computer, open the internet and browse around the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['beer', 'toilet', 'chair', 'walllamp', 'blow_dryer', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'food_orange', 'wall', 'powersocket', 'cup', 'standingmirror', 'keys', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "turn computer on, open application to work with", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'detergent', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'thread', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'photoframe', 'toilet_paper', 'television', 'razor', 'coin', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Push power button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'towel', 'wall', 'wall_clock', 'oven', 'television', 'tray', 'rag', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'laser_pointer']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <towel>", "[GRAB] <towel>", "[CLOSE] <bathroom_cabinet>", "[FIND] <shower>", "[RINSE] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'drawing', 'wallshelf', 'food_egg', 'hanger', 'cpuscreen', 'floor', 'ceiling', 'character', 'keyboard', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'toaster', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'bed', 'television', 'curtain', 'doorjamb', 'dresser', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room and turn on the light. I walk to the couch then sit on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'microwave', 'toilet', 'chair', 'blow_dryer', 'hanger', 'bathroom_counter', 'ceiling', 'phone', 'desk', 'mail', 'shoe-shine_kit', 'mousepad', 'food_noodles', 'cutting_board', 'cup', 'photoframe', 'bed', 'coffe_maker', 'television', 'curtain', 'centerpiece', 'comb', 'pajamas', 'clothes_dress', 'door', 'home_office', 'light', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room, sit down by the cat, put out hand, stroke the cat from head to tail", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'orchid', 'drawing', 'floor', 'ceiling', 'toaster', 'food_noodles', 'mousepad', 'food_peanut_butter', 'wall', 'oven', 'cup', 'bed', 'photoframe', 'coffe_maker', 'toilet_paper', 'television', 'coffee_filter', 'ice', 'curtain', 'console', 'kitchen_counter', 'cupboard', 'sink', 'clothes_scarf', 'dishwasher', 'soap', 'dirt', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'freezer', 'bathroom', 'detergent', 'bathroom_counter', 'floor', 'bedroom', 'toaster', 'plate', 'shower', 'dining_room', 'nightstand', 'oven', 'cup', 'coffe_maker', 'food_food', 'television', 'tray', 'kitchen_counter', 'dresser', 'door', 'sink', 'bathtub', 'home_office', 'table', 'dishwasher', 'soap', 'board_game']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'toilet', 'jelly', 'freezer', 'ground_coffee', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'button', 'mat', 'shower', 'picture', 'dustpan', 'food_orange', 'form', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'light_bulb', 'food_food', 'doorjamb', 'cd', 'door', 'light']\nTask:Added meat to freezer\nresponse:\n{\n    \"thoughts\": After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'scissors', 'knife', 'hanger', 'bathroom_counter', 'floor', 'bedroom', 'desk', 'keyboard', 'piano_bench', 'stovefan', 'stereo', 'window', 'wall', 'powersocket', 'cutting_board', 'cup', 'bed', 'console', 'pencil', 'shoes', 'door', 'bathtub', 'bag', 'instrument_guitar', 'table', 'home_office', 'light', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into my room and turn the radio on so that I can listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'character', 'desk', 'bookshelf', 'mat', 'deck_of_cards', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'bed', 'coffee_filter', 'wine', 'curtain', 'doorjamb', 'kitchen_counter', 'pajamas', 'door', 'food_steak', 'bag', 'light', 'electrical_outlet']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the electric outlet. I switch on the light.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <electrical_outlet>', '[FIND] <electrical_outlet>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'beer', 'crayon', 'walllamp', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'bookshelf', 'mat', 'iron', 'electric_shaver', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'toy', 'cup', 'light_bulb', 'novel', 'food_food', 'check', 'toothbrush_holder', 'television', 'curtain', 'doorjamb', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I go to the book case, I select a book, I go back and sit down, then I read.", "actions": ["[WALK] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'dishwasher']\nTask:Work\nresponse:\n{\n    \"thoughts\": Sit at my computer desk, Turn on Computer. Get online. Work.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['chair', 'toothbrush', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'computer', 'table_cloth', 'toaster', 'laptop', 'shower', 'mousepad', 'closetdrawer', 'dustpan', 'dining_room', 'wall', 'powersocket', 'food_cheese', 'coffe_maker', 'check', 'wooden_spoon', 'doorjamb', 'mouse', 'dresser', 'door', 'maindoor', 'food_steak', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "first ill go to my home office adn ill switch on my laptop and with the use of mouse and keyboard il run my computer", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[RUN] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'milk', 'toothbrush', 'drawing', 'orchid', 'cpuscreen', 'character', 'bedroom', 'stovefan', 'toaster', 'mousepad', 'clothes_pants', 'food_kiwi', 'wall_clock', 'bed', 'television', 'toothbrush_holder', 'water', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'shoes', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'dirt']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <toothbrush>']\n}\n", "input": "\nScene_Objects:['cat', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'iron', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'tape', 'novel', 'food_food', 'television', 'tray', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[PUTOBJBACK] <soap>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'knife', 'cpuscreen', 'hanger', 'floor', 'phone', 'desk', 'keyboard', 'thread', 'computer', 'piano_bench', 'toaster', 'laptop', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'box', 'console', 'mouse', 'dresser', 'maindoor', 'hairdryer', 'board_game']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I open my laptop and then login to Windows. Then I open Google Chrome, and type gmail.com into the address bar. My email automatically logs in, then I click the Compose button, type in the address of the recipient, then the subject of the email, and then start typing my email,\n    \"actions\": ['[WALK] <laptop>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'computer', 'button', 'bookmark', 'mat', 'picture', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'novel', 'check', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'light', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Find book, grab book, fin bookmark, sit down somewhere comfortable, open book, read book, close book after reading.", "actions": ["[WALK] <desk>", "[FIND] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[FIND] <couch>", "[SIT] <couch>", "[GRAB] <novel>", "[READ] <novel>", "[PUTBACK] <bookmark> <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'scissors', 'chair', 'hanger', 'phone', 'keyboard', 'bookshelf', 'plate', 'faucet', 'window', 'mousepad', 'food_noodles', 'oil', 'food_peanut_butter', 'clothes_pants', 'dining_room', 'bench', 'oven', 'cutting_board', 'toy', 'basket_for_clothes', 'cup', 'photoframe', 'keys', 'food_food', 'laundry_detergent', 'bathroom_cabinet', 'instrument_violin', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'jelly', 'drawing', 'floor', 'bedroom', 'table_cloth', 'stovefan', 'mat', 'plate', 'window', 'couch', 'wall', 'nightstand', 'wall_clock', 'cutting_board', 'bed', 'novel', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'pencil', 'cd', 'clothes_dress', 'bathroom_cabinet', 'table', 'light', 'vase', 'laser_pointer']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I would go lie down in my bed and open the book and start reading.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'crayon', 'dog', 'freezer', 'knife', 'orchid', 'bathroom_counter', 'character', 'bedroom', 'faucet', 'couch', 'food_orange', 'towel', 'nightstand', 'oven', 'powersocket', 'photoframe', 'check', 'video_game_controller', 'food_donut', 'food_rice', 'food_onion', 'diary', 'dresser', 'hairbrush', 'clothes_scarf', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the cat. Pet the cat with your hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'dog', 'cat', 'dough', 'clothes_underwear', 'folder', 'closetdrawer', 'couch', 'food_orange', 'broom', 'oven', 'powersocket', 'toy', 'photoframe', 'bed', 'check', 'razor', 'curtain', 'lighter', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'door', 'sink', 'clothes_scarf', 'napkin', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'trashcan', 'hanger', 'floor', 'bedroom', 'computer', 'button', 'brush', 'mat', 'shower', 'faucet', 'window', 'clothes_hat', 'wall', 'child', 'oven', 'bed', 'newspaper', 'novel', 'curtain', 'centerpiece', 'doorjamb', 'mouse', 'cd', 'dresser', 'hairdryer', 'clothes_scarf', 'vase', 'board_game']\nTask:Story reading time\nresponse:\n{\n    \"thoughts\": Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[TURNTO] <child>', '[LOOKAT] <child>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'mop', 'toilet', 'chair', 'walllamp', 'bathroom', 'food_carrot', 'hanger', 'floor', 'shampoo', 'bookshelf', 'purse', 'computer', 'stovefan', 'toaster', 'shower', 'mousepad', 'dustpan', 'clothes_pants', 'wall', 'cutting_board', 'standingmirror', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'hairbrush', 'door', 'bathtub']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom. Pull down pants. Use the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[PULL] <clothes_pants>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'mop_bucket', 'stovefan', 'ceilinglamp', 'tvstand', 'homework', 'mousepad', 'couch', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'television', 'doorjamb', 'mouse', 'headset', 'cupboard', 'table', 'shoe_rack']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I go to the home office to find the computer. I turn switch computer on , i open the email. i write the email and  send it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'toilet', 'chair', 'walllamp', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'shower', 'mousepad', 'food_peanut_butter', 'food_jam', 'wall', 'bench', 'powersocket', 'toy', 'box', 'standingmirror', 'coffe_maker', 'water', 'doorjamb', 'cards', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'light', 'soap', 'board_game']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "go to a bathroom, open the shower, let water wash your body", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'toilet', 'walllamp', 'cat', 'orchid', 'floor', 'ceiling', 'remote_control', 'computer', 'food_bacon', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'mouthwash', 'television', 'coin', 'food_rice', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'table', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'band-aids', 'dough', 'food_carrot', 'drawing', 'bookshelf', 'stovefan', 'shower', 'oil', 'towel', 'dining_room', 'wall', 'bench', 'toy', 'shelf', 'box', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'food_food', 'food_rice', 'kettle', 'shoes', 'door', 'bathtub', 'home_office', 'vase', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "BEFORE EATING", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <towel>", "[WIPE] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'walllamp', 'food_carrot', 'knife', 'orchid', 'cpuscreen', 'phone', 'dvd_player', 'desk', 'mop_bucket', 'toaster', 'laptop', 'tvstand', 'closetdrawer', 'food_jam', 'pillow', 'dining_room', 'food_food', 'television', 'oven_mitts', 'water', 'after_shave', 'pencil', 'hairbrush', 'clothes_dress', 'door', 'table', 'dishwasher', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['walllamp', 'toothbrush', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'thread', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'plate', 'dining_room', 'wall', 'wall_clock', 'oven', 'toothbrush_holder', 'tray', 'doorjamb', 'towel_rack', 'chessboard', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "getting ready for bed", "actions": ["[WALK] <bathroom_counter>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <toothbrush> <bathroom_counter>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[GRAB] <toothbrush>", "[SCRUB] <toothbrush>", "[RINSE] <toothbrush>", "[FIND] <toothbrush_holder>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'trashcan', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'thread', 'computer', 'piano_bench', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'oven', 'bed', 'novel', 'check', 'television', 'curtain', 'towel_rack', 'dresser', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Put child in bed, sit down,read book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'glue', 'food_ice_cream', 'food_salt', 'tea', 'drawing', 'knife', 'cpuscreen', 'phone', 'desk', 'computer', 'button', 'plate', 'iron', 'dish_soap', 'mousepad', 'nightstand', 'powersocket', 'curtain', 'kitchen_counter', 'dishrack', 'dresser', 'spoon', 'napkin', 'table', 'dishwasher', 'home_office', 'vase', 'bathroom_cabinet', 'coffee']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <dishrack>", "[PUTBACK] <fork> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishrack>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishrack>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'jelly', 'ceilingfan', 'food_carrot', 'drawing', 'knife', 'clothes_underwear', 'floor', 'phone', 'ceilinglamp', 'tvstand', 'stereo', 'window', 'mousepad', 'couch', 'food_jam', 'wall', 'nightstand', 'cup', 'television', 'ice', 'mouse', 'kitchen_counter', 'home_office', 'table', 'soap', 'board_game', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": sitting down to listen to music on the radio,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'deck_of_cards', 'window', 'homework', 'closetdrawer', 'couch', 'pillow', 'wall', 'standingmirror', 'newspaper', 'photoframe', 'food_food', 'television', 'food_donut', 'curtain', 'doorjamb', 'towel_rack', 'bag', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Lift the remote and change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'food_bacon', 'alcohol', 'ceilinglamp', 'pillow', 'wall', 'powersocket', 'bed', 'curtain', 'doorjamb', 'kettle', 'mouse', 'dresser', 'cupboard', 'door', 'maindoor', 'light', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I open the room door and press the light button on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['bathroom', 'detergent', 'milk', 'floor', 'folder', 'bedroom', 'tea_bag', 'remote_control', 'button', 'mat', 'plate', 'shower', 'closetdrawer', 'pillow', 'nightstand', 'toy', 'standingmirror', 'clothes_gloves', 'keys', 'food_food', 'tray', 'after_shave', 'towel_rack', 'shoes', 'maindoor', 'bathtub', 'sink', 'light', 'table', 'scrabble']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk into living room. Find television remote control. Pick up remote control. Press button.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'blender', 'crayon', 'microwave', 'cat', 'drawing', 'orchid', 'needle', 'phone', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'food_jam', 'dining_room', 'powersocket', 'bed', 'photoframe', 'coffe_maker', 'television', 'diary', 'shoes', 'comb', 'controller', 'food_steak', 'instrument_violin', 'sink', 'dishwasher', 'board_game']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the cat. I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'band-aids', 'freezer', 'orchid', 'drawing', 'cpuscreen', 'bills', 'floor', 'ceiling', 'character', 'bookshelf', 'mop_bucket', 'mat', 'deck_of_cards', 'faucet', 'window', 'electric_shaver', 'couch', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'coffee_filter', 'curtain', 'doorjamb', 'mouse', 'conditioner', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room ,find sofa ,sit on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'jelly', 'band-aids', 'chair', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'shampoo', 'ceiling', 'desk', 'keyboard', 'purse', 'brush', 'toaster', 'shoe-shine_kit', 'plate', 'faucet', 'mousepad', 'towel', 'dining_room', 'wall', 'photoframe', 'newspaper', 'coffe_maker', 'toothbrush_holder', 'mouse', 'door', 'home_office']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. I walk over to the toilet. I sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['beer', 'tea', 'chair', 'walllamp', 'ground_coffee', 'bills', 'orchid', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'mail', 'ceilinglamp', 'mat', 'iron', 'stereo', 'faucet', 'clothes_socks', 'clothes_hat', 'basket_for_clothes', 'food_rice', 'curtain', 'lighter', 'doorjamb', 'towel_rack', 'vacuum_cleaner', 'dishrack', 'cd', 'napkin', 'home_office']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TOUCH] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['knife', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'folder', 'desk', 'keyboard', 'computer', 'mat', 'stereo', 'mousepad', 'closetdrawer', 'clothes_pants', 'wall', 'powersocket', 'cup', 'newspaper', 'novel', 'tray', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'spoon', 'light', 'soap', 'board_game', 'clothes_shirt', 'scrabble']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Pick out the clothes you are going to wear and put them on your body,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'walllamp', 'blow_dryer', 'milk', 'bills', 'orchid', 'floor', 'ceiling', 'character', 'food_cake', 'alcohol', 'stovefan', 'homework', 'closetdrawer', 'pillow', 'broom', 'wall', 'oven', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'kitchen_counter', 'shoes', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <freezer>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <cup>", "[POUR] <water> <cup>", "[FIND] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'mop', 'chair', 'bathroom', 'drawing', 'wallshelf', 'character', 'bedroom', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'shower', 'faucet', 'floor_lamp', 'mousepad', 'closetdrawer', 'nightstand', 'wall', 'oven', 'bench', 'photoframe', 'bed', 'coffe_maker', 'food_food', 'check', 'conditioner', 'light', 'table', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to living room, find lamp, switch on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['jelly', 'freezer', 'dog', 'chair', 'nail_polish', 'bathroom', 'orchid', 'bills', 'food_egg', 'hanger', 'character', 'bedroom', 'thread', 'toaster', 'faucet', 'mousepad', 'oil', 'nightstand', 'cup', 'photoframe', 'standingmirror', 'bed', 'video_game_controller', 'wine', 'doorjamb', 'maindoor', 'bathtub', 'vase', 'clothes_shirt', 'coffee']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up  the book open the book and read", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'towel', 'wall', 'wall_clock', 'oven', 'television', 'tray', 'rag', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <bathroom_cabinet>', '[FIND] <shower>', '[RINSE] <shower>']\n}\n", "input": "\nScene_Objects:['scissors', 'toilet', 'freezer', 'bathroom', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'thread', 'stovefan', 'mat', 'faucet', 'towel', 'dining_room', 'wall', 'wall_clock', 'oven', 'bench', 'nightstand', 'check', 'coffee_filter', 'tray', 'shaving_cream', 'towel_rack', 'pencil', 'mouse', 'sink', 'bathroom_cabinet', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. Walk to the sink. Turn on tap. Get soap. Wash and rinse hands. Turn off tap. Dry hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[LOOKAT] <sink>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[RINSE] <soap>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'ceilinglamp', 'mat', 'shower', 'food_noodles', 'couch', 'towel', 'wall', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'newspaper', 'water', 'food_donut', 'curtain', 'doorjamb', 'headset', 'door', 'bathroom_cabinet', 'light', 'soap', 'tooth_paste']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>', '[TURNTO] <shower>', '[FIND] <water>', '[RINSE] <water>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <water>', '[FIND] <towel>', '[GRAB] <towel>', '[SCRUB] <towel>']\n}\n", "input": "\nScene_Objects:['scissors', 'dog', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'remote_control', 'envelope', 'computer', 'table_cloth', 'mat', 'ceilinglamp', 'window', 'food_noodles', 'closetdrawer', 'couch', 'pillow', 'wall', 'bench', 'standingmirror', 'television', 'tray', 'curtain', 'doorjamb', 'console', 'cd', 'napkin', 'table', 'clothes_shirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the television. Walk to the television. Switch on the television. Find the sofa. Walk to the sofa. Sit on the sofa. Find the remote control. Grab the remote control. Look at the television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[WALK] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_salt', 'beer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'bookshelf', 'mail', 'computer', 'button', 'toaster', 'plate', 'shower', 'dish_soap', 'mousepad', 'pillow', 'wall', 'oven', 'photoframe', 'clothes_gloves', 'coffe_maker', 'tray', 'food_onion', 'doorjamb', 'dresser', 'hairdryer', 'sink', 'dishwasher', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put dishes in the dishwasher and start the dishwasher.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[WALK] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'floor', 'ceiling', 'keyboard', 'stovefan', 'electric_shaver', 'clothes_socks', 'form', 'wall', 'nightstand', 'oven', 'cutting_board', 'coffe_maker', 'food_food', 'oven_mitts', 'food_donut', 'tray', 'doorjamb', 'kettle', 'mouse', 'kitchen_counter', 'laundry_detergent', 'home_office', 'table', 'bathroom_cabinet', 'instrument_guitar']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "go to the kitchen and put the groceries in the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'filing_cabinet', 'scissors', 'chair', 'ceilingfan', 'dough', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'coffee_table', 'television', 'coin', 'wooden_spoon', 'curtain', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'envelope', 'computer', 'mop_bucket', 'stovefan', 'mat', 'tvstand', 'toaster', 'foundation', 'mousepad', 'pillow', 'wall', 'standingmirror', 'photoframe', 'toilet_paper', 'oven_mitts', 'tray', 'mouse', 'sink']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to living room,find the text book on the self ,sat on the chair ,open and read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'computer', 'mail', 'table_cloth', 'mat', 'tvstand', 'plate', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'curtain', 'after_shave', 'doorjamb', 'cards', 'mouse', 'shoes', 'sink']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I sit in the sofa and I relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['crayon', 'walllamp', 'blow_dryer', 'detergent', 'orchid', 'bathroom_counter', 'ceiling', 'phone', 'computer', 'clothes_jacket', 'table_cloth', 'stovefan', 'faucet', 'shower', 'couch', 'clothes_pants', 'wall', 'wall_clock', 'oven', 'bench', 'bed', 'toothbrush_holder', 'tray', 'curtain', 'dresser', 'clothes_dress', 'sink', 'spoon', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open cabinet and collect clothes and put them into washing machine and wash clothes", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'table']\nTask:Watch youtube\nresponse:\n{\n    \"thoughts\": Go to youtube.Com watch videos,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['dog', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'stereo', 'mousepad', 'pillow', 'towel', 'wall', 'nightstand', 'bench', 'cutting_board', 'cup', 'food_food', 'coffe_maker', 'television', 'food_donut', 'coffee_pot', 'towel_rack', 'soap']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk into living room, walk to stereo, turn on stereo, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'dog', 'freezer', 'walllamp', 'cat', 'knifeblock', 'bills', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'faucet', 'clothes_hat', 'broom', 'powersocket', 'bench', 'coffe_maker', 'food_food', 'television', 'mirror', 'rag', 'lighter', 'towel_rack', 'mouse', 'dresser', 'sink', 'table', 'clothes_skirt', 'shoe_rack']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": Spray glass cleaner on mirror. Wipe with a clean cloth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[WALK] <mirror>', '[PUTOBJBACK] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'blender', 'beer', 'toilet', 'detergent', 'food_butter', 'phone', 'bookshelf', 'remote_control', 'alcohol', 'button', 'table_cloth', 'piano_bench', 'toaster', 'tvstand', 'pillow', 'nightstand', 'pasta', 'wall_clock', 'tape', 'coffe_maker', 'doorjamb', 'cards', 'pencil', 'controller', 'clothes_dress', 'maindoor', 'sink', 'home_office', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pick up the remote. I press button on remote.", "actions": ["[WALK] <bedroom>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'scissors', 'tea', 'freezer', 'walllamp', 'knifeblock', 'knife', 'bills', 'floor', 'toaster', 'mat', 'plate', 'stereo', 'picture', 'tvstand', 'dish_soap', 'microphone', 'wall', 'oven', 'cup', 'light_bulb', 'coffe_maker', 'drinking_glass', 'tray', 'mouse', 'cupboard', 'controller', 'maindoor', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen and turn to the dishwasher, put the knife, glass, fork and plate into the dishwasher. Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[TURNTO] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[OPEN] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'freezer', 'band-aids', 'walllamp', 'food_carrot', 'drawing', 'floor', 'ceiling', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'wall', 'nightstand', 'oven', 'cutting_board', 'cup', 'food_cheese', 'coffe_maker', 'coffee_filter', 'drinking_glass', 'tray', 'shaving_cream', 'kitchen_counter', 'dresser', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Robot Drinks the water", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>", "[PUTOBJBACK] <drinking_glass>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['cd_player', 'freezer', 'ground_coffee', 'bathroom', 'milk', 'orchid', 'cpuscreen', 'hanger', 'dvd_player', 'character', 'ceilinglamp', 'stereo', 'faucet', 'dustpan', 'clothes_hat', 'nightstand', 'wall', 'newspaper', 'check', 'mouthwash', 'video_game_controller', 'food_food', 'tray', 'doorjamb', 'pencil', 'cupboard', 'door', 'sink', 'home_office', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk in the living room, then I go the the stereo and the cd player and turn them on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'food_egg', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'bookshelf', 'piano_bench', 'shower', 'faucet', 'stereo', 'electric_shaver', 'food_jam', 'wall', 'nightstand', 'wall_clock', 'shelf', 'bed', 'photoframe', 'novel', 'light_bulb', 'coffee_filter', 'board_game', 'laser_pointer', 'shoe_rack']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Travel through the house and locate the bathroom, and proceed to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'orchid', 'ceiling', 'desk', 'computer', 'food_bacon', 'stovefan', 'ceilinglamp', 'plate', 'shower', 'faucet', 'window', 'form', 'dining_room', 'oven', 'basket_for_clothes', 'keys', 'novel', 'food_food', 'coffe_maker', 'food_rice', 'tray', 'doorjamb', 'cards', 'mouse', 'pajamas', 'home_office', 'light', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the playback device you want to listen to. turn on. press play. enjoyu music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['fork', 'food_salt', 'food_butter', 'food_carrot', 'drawing', 'food_egg', 'floor', 'phone', 'ceiling', 'character', 'purse', 'mat', 'plate', 'window', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'bed', 'photoframe', 'tray', 'coffee_pot', 'kitchen_counter', 'shoes', 'pajamas', 'sink', 'spoon', 'bathtub', 'table', 'vase']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Place your plate on the placemat. The forks will be placed on the left side of the plate and the knife and spoon on the right and place the spoon to the right of the knife. If you want to set out food, place them above the plate.", "actions": ["[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[DROP] <plate>", "[FIND] <fork>", "[GRAB] <fork>", "[DROP] <fork>", "[FIND] <spoon>", "[GRAB] <spoon>", "[DROP] <spoon>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'bookmark', 'ceilinglamp', 'tvstand', 'balanceball', 'picture', 'mousepad', 'couch', 'wall', 'bed', 'food_food', 'tray', 'after_shave', 'rag', 'doorjamb', 'maindoor', 'bathroom_cabinet', 'dishwasher', 'vase']\nTask:Put away groceries\nresponse:\n{\n    \"thoughts\": Place groceries in refrigerator from shopping at the grocery store.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'bedroom', 'mail', 'stovefan', 'faucet', 'stereo', 'couch', 'wall', 'oven', 'cutting_board', 'bed', 'newspaper', 'coffe_maker', 'food_food', 'tray', 'lighter', 'doorjamb', 'kitchen_counter', 'home_office', 'bathroom_cabinet', 'vase', 'soap']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Open fridge. Put groceries in fridge.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'food_butter', 'knifeblock', 'floor', 'ceiling', 'character', 'desk', 'food_cake', 'computer', 'bookmark', 'toaster', 'stovefan', 'mat', 'faucet', 'food_kiwi', 'wall', 'oven', 'powersocket', 'food_vegetable', 'bed', 'food_food', 'oven_mitts', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I prepare meals and cook them on the stove.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[CLOSE] <freezer>', '[FIND] <oven>', '[SWITCHON] <oven>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'freezer', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stovefan', 'mat', 'tvstand', 'faucet', 'toaster', 'teeth', 'towel', 'wall', 'wall_clock', 'oven', 'powersocket', 'food_food', 'tray', 'headset', 'comb', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'instrument_guitar', 'soap', 'hands_both', 'creditcard']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to bathroom, turn on faucet, wet hands, apply soap, rub hands, rinse hands, turn off faucet, dry hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'walllamp', 'bathroom', 'knifeblock', 'food_egg', 'food_bacon', 'brush', 'mat', 'deck_of_cards', 'balanceball', 'tvstand', 'ceilinglamp', 'picture', 'couch', 'wall', 'tape', 'cup', 'toilet_paper', 'food_food', 'video_game_controller', 'water', 'cupboard', 'bathroom_cabinet', 'table', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_salt', 'microwave', 'freezer', 'walllamp', 'floor', 'ceiling', 'bedroom', 'food_cake', 'mail', 'table_cloth', 'stovefan', 'mat', 'shower', 'closetdrawer', 'food_jam', 'wall', 'nightstand', 'oven', 'cutting_board', 'shelf', 'bed', 'photoframe', 'standingmirror', 'coffe_maker', 'food_food', 'tray', 'kitchen_counter', 'dresser', 'instrument_violin']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Put groceries into the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'walllamp', 'nail_polish', 'dough', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'cards', 'dishrack', 'door', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'food_ice_cream', 'crayon', 'toilet', 'chair', 'freezer', 'jelly', 'detergent', 'floor', 'ceiling', 'bookshelf', 'table_cloth', 'deck_of_cards', 'shower', 'picture', 'couch', 'wall', 'shelf', 'newspaper', 'wooden_spoon', 'centerpiece', 'kettle', 'shoes', 'comb', 'conditioner', 'door', 'napkin', 'bathroom_cabinet', 'table']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "i go to the bathroom. I find the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'blow_dryer', 'cat', 'trashcan', 'toothbrush', 'drawing', 'floor', 'character', 'desk', 'wine_glass', 'ceilinglamp', 'plate', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'food_cheese', 'bed', 'food_food', 'check', 'television', 'coffee_filter', 'curtain', 'door', 'spoon', 'table']\nTask:Entertain\nresponse:\n{\n    \"thoughts\": Give guests wine, cheese, crackers and sit and talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <food_cheese>', '[FIND] <food_cheese>', '[FIND] <food_food>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <wine_glass>', '[GRAB] <wine_glass>', '[DRINK] <wine_glass>']\n}\n", "input": "\nScene_Objects:['mop', 'toilet', 'freezer', 'food_carrot', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'piano_bench', 'stovefan', 'mat', 'faucet', 'wall', 'wall_clock', 'oven', 'cutting_board', 'toilet_paper', 'video_game_controller', 'television', 'tray', 'doorjamb', 'towel_rack', 'coffee_pot', 'kitchen_counter', 'dresser', 'conditioner', 'sink', 'bathroom_cabinet', 'light', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom.  Find mouthwash or turn on water.  Put mouthwash or water in mouth.  Swish around for 30 seconds.  Spit mouthwash out.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <teeth>", "[RINSE] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'blow_dryer', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'stovefan', 'ceilinglamp', 'balanceball', 'stereo', 'window', 'shower', 'clothes_socks', 'couch', 'candle', 'broom', 'wall', 'wall_clock', 'bed', 'television', 'curtain', 'pencil', 'kitchen_counter', 'dresser', 'napkin', 'spoon', 'table', 'dirt', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Find the sofa and sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'microwave', 'toilet', 'chair', 'dog', 'walllamp', 'floor', 'desk', 'bookshelf', 'keyboard', 'mat', 'plate', 'shower', 'window', 'bowl', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'shelf', 'check', 'food_food', 'cards', 'towel_rack', 'kitchen_counter', 'maindoor', 'table', 'vase', 'dirt']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[OPEN] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'scissors', 'mop', 'chair', 'ceilingfan', 'milk', 'toothbrush', 'knifeblock', 'phone', 'bookshelf', 'stovefan', 'toaster', 'tvstand', 'shower', 'couch', 'food_peanut_butter', 'dustpan', 'oven', 'photoframe', 'clothes_gloves', 'bed', 'video_game_controller', 'tray', 'towel_rack', 'clothes_scarf', 'table', 'light', 'soap', 'tooth_paste', 'scrabble']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": before going to bed or when wake up earily morning brush teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'shelf', 'cup', 'television', 'doorjamb', 'mouse', 'address_book', 'dresser', 'conditioner', 'bathtub', 'spoon', 'light', 'table', 'soap']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office. Walk to computer. Switch on computer. Open the address book.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <address_book>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'mop', 'cat', 'knife', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'dvd_player', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'faucet', 'stereo', 'window', 'couch', 'towel', 'wall', 'novel', 'tray', 'doorjamb', 'hairbrush', 'sink', 'bathroom_cabinet', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'freezer', 'orchid', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'phone', 'bookshelf', 'stovefan', 'pillow', 'wall', 'wall_clock', 'cutting_board', 'box', 'shelf', 'newspaper', 'bed', 'novel', 'coffe_maker', 'toilet_paper', 'wooden_spoon', 'hairbrush', 'door', 'home_office', 'table', 'light', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up the book, open to current page, read the words.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'freezer', 'dough', 'milk', 'drawing', 'floor', 'ceiling', 'keyboard', 'food_bacon', 'toaster', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'bowl', 'wall', 'oven', 'cup', 'coffe_maker', 'check', 'television', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'light', 'dishwasher']\nTask:Get something to drink\nresponse:\n{\n    \"thoughts\": Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[CLOSE] <freezer>', '[POUR] <milk> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'drawing', 'needle', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'dustpan', 'dining_room', 'wall', 'nightstand', 'powersocket', 'food_food', 'toilet_paper', 'doorjamb', 'coffee_pot', 'mouse', 'kettle', 'dresser', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk into living room, walk to light, turn on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'ground_coffee', 'food_butter', 'toothbrush', 'knifeblock', 'drawing', 'phone', 'keyboard', 'bookshelf', 'table_cloth', 'mat', 'plate', 'electric_shaver', 'window', 'pillow', 'nightstand', 'oven', 'newspaper', 'clothes_gloves', 'photoframe', 'toilet_paper', 'water', 'centerpiece', 'towel_rack', 'kettle', 'pencil', 'clothes_scarf', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[FIND] <toothbrush>', '[FIND] <water>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['glue', 'microwave', 'freezer', 'ground_coffee', 'hanger', 'floor', 'ceiling', 'stamp', 'keyboard', 'toaster', 'plate', 'faucet', 'shower', 'mousepad', 'food_kiwi', 'wall', 'wall_clock', 'oven', 'cutting_board', 'standingmirror', 'photoframe', 'coffe_maker', 'video_game_controller', 'razor', 'coffee_filter', 'water', 'coffee_pot', 'kitchen_counter', 'sink', 'home_office']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'drawing', 'floor', 'purse', 'computer', 'plate', 'balanceball', 'shower', 'stereo', 'dustpan', 'clothes_hat', 'microphone', 'bench', 'basket_for_clothes', 'cup', 'bed', 'check', 'coffe_maker', 'food_onion', 'after_shave', 'cards', 'pencil', 'chessboard', 'maindoor', 'napkin', 'table', 'board_game', 'clothes_shirt', 'coffee']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'bathroom', 'cat', 'milk', 'orchid', 'floor', 'dvd_player', 'desk', 'keyboard', 'purse', 'table_cloth', 'mat', 'shower', 'window', 'clothes_socks', 'couch', 'clothes_pants', 'pillow', 'wall', 'powersocket', 'mouthwash', 'television', 'curtain', 'food_onion', 'doorjamb', 'kitchen_counter', 'bathtub', 'napkin', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'walllamp', 'drawing', 'hanger', 'floor', 'mat', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'cutting_board', 'cup', 'novel', 'food_food', 'television', 'wine', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'shoes', 'bathtub', 'bathroom_cabinet', 'table', 'soap', 'board_game']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": lie in the sofa, relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'toilet', 'walllamp', 'nail_polish', 'food_butter', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'ceilinglamp', 'tvstand', 'shower', 'closetdrawer', 'couch', 'form', 'dining_room', 'wall', 'wall_clock', 'bed', 'toothbrush_holder', 'doorjamb', 'chessboard', 'door', 'napkin', 'bathroom_cabinet', 'shoe_rack']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone and pick it up.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'microwave', 'detergent', 'knifeblock', 'floor', 'ceiling', 'character', 'computer', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'wall', 'cup', 'check', 'mouthwash', 'water', 'wine', 'after_shave', 'curtain', 'food_onion', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'conditioner', 'sink', 'napkin', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'table_cloth', 'ceilinglamp', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'shelf', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office', 'light', 'table', 'laser_pointer']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Find the computer.  Turn on the computer.  Open the e-mail program.  Create a new message.  Write the message.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'scissors', 'juice', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'faucet', 'towel', 'dining_room', 'wall', 'photoframe', 'food_food', 'razor', 'television', 'food_rice', 'wooden_spoon', 'after_shave', 'headset', 'laundry_detergent', 'conditioner', 'sink', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Turn on the water faucet. Wet my hands and turn off the faucet. Put soap on my hands. Rub hands together, making sure the soap covers all areas of the hands. Scrub hands for about 20 seconds. Turn on the water water faucet. Rinse hands under the running water. Turn off the water faucet. Dry hands with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'microwave', 'chair', 'freezer', 'nail_polish', 'detergent', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'food_cake', 'keyboard', 'toaster', 'stovefan', 'window', 'tooth_paste', 'cup', 'check', 'toothbrush_holder', 'doorjamb', 'coffee_pot', 'mouse', 'pencil', 'bathtub', 'light', 'table', 'board_game', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go looking for the cat in the living room where he usually sleeps. Once I find the cat I just pet him for a few minutes.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'bathroom', 'orchid', 'knifeblock', 'cpuscreen', 'needle', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'mousepad', 'couch', 'dining_room', 'wall', 'basket_for_clothes', 'light_bulb', 'food_food', 'check', 'toothbrush_holder', 'mouse', 'comb', 'door', 'clothes_scarf', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk to the home office. I open the door of the office room and sit on the chair and in front of me is my desktop computer,  and I switch on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'cat', 'floor', 'phone', 'dvd_player', 'bedroom', 'stovefan', 'mat', 'window', 'electric_shaver', 'couch', 'pillow', 'towel', 'wall', 'powersocket', 'toy', 'food_cheese', 'photoframe', 'newspaper', 'television', 'curtain', 'doorjamb', 'pencil', 'shoes', 'maindoor', 'bathtub', 'spoon', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'microwave', 'toilet', 'cat', 'wallshelf', 'bathroom_counter', 'floor', 'dvd_player', 'bookshelf', 'stovefan', 'ceilinglamp', 'toaster', 'mat', 'shower', 'faucet', 'tvstand', 'couch', 'towel', 'wall', 'bench', 'cutting_board', 'bed', 'razor', 'television', 'curtain', 'kitchen_counter', 'dresser', 'home_office', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Pet Cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'microwave', 'freezer', 'walllamp', 'detergent', 'toothbrush', 'orchid', 'drawing', 'clothes_underwear', 'floor', 'thread', 'mat', 'closetdrawer', 'wall', 'nightstand', 'toy', 'standingmirror', 'light_bulb', 'novel', 'check', 'tray', 'cards', 'mouse', 'pencil', 'vacuum_cleaner', 'headset', 'maindoor', 'napkin', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to the Living Room. Look at the Cabinet. Find a Textbook. Grab the textbook. Open the Textbook. Read the textbook.", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[FIND] <filing_cabinet>", "[TURNTO] <filing_cabinet>", "[LOOKAT] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'trashcan', 'food_carrot', 'knifeblock', 'drawing', 'cpuscreen', 'clothes_underwear', 'floor', 'bedroom', 'alcohol', 'laptop', 'tvstand', 'stereo', 'faucet', 'mat', 'couch', 'clothes_hat', 'nightstand', 'bench', 'oven', 'toy', 'coffe_maker', 'television', 'coin', 'kitchen_counter', 'cupboard', 'home_office', 'bathroom_cabinet', 'dishwasher', 'board_game']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": go to radio. run on radio.,\n    \"actions\": ['[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['scissors', 'detergent', 'orchid', 'wallshelf', 'bedroom', 'tea_bag', 'bookshelf', 'ceilinglamp', 'laptop', 'faucet', 'window', 'couch', 'broom', 'wall', 'oven', 'cutting_board', 'toy', 'cup', 'standingmirror', 'razor', 'television', 'tray', 'after_shave', 'towel_rack', 'mouse', 'kitchen_counter', 'controller', 'bathtub', 'vase', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room, walk to the phone reach for the phone and pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'food_salt', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'desk', 'alcohol', 'stovefan', 'mat', 'plate', 'shower', 'couch', 'food_orange', 'pillow', 'wall', 'oven', 'food_food', 'television', 'tray', 'towel_rack', 'kitchen_counter', 'sauce_pan', 'cupboard', 'spoon', 'napkin', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <spoon>', '[GRAB] <spoon>', '[FIND] <sauce_pan>', '[PUTBACK] <spoon> <sauce_pan>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <plate>', '[PUTBACK] <pot> <plate>']\n}\n", "input": "\nScene_Objects:['spectacles', 'fork', 'tablelamp', 'tea', 'band-aids', 'bathroom', 'drawing', 'hanger', 'floor', 'ceiling', 'dvd_player', 'desk', 'remote_control', 'computer', 'food_bacon', 'mat', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'clothes_pants', 'pillow', 'wall', 'standingmirror', 'television', 'curtain', 'doorjamb', 'table', 'soap']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I push my body to be near to the table in front of me, without stand up. I move my right hand to near the TV remote control on the table. I take the TV remote control. I look the number on the remote control. I touch number 3 button. I put back the remote control to the table. I put my body back to the sofa", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[GRAB] <remote_control>", "[TOUCH] <remote_control>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'detergent', 'food_carrot', 'orchid', 'floor', 'wine_glass', 'food_cake', 'mat', 'picture', 'window', 'mousepad', 'couch', 'clothes_hat', 'pillow', 'wall', 'pasta', 'shelf', 'food_cheese', 'newspaper', 'food_food', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'bathtub', 'table', 'vase']\nTask:Entertain\nresponse:\n{\n    \"thoughts\": Give guests wine, cheese, crackers and sit and talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <food_cheese>', '[FIND] <food_cheese>', '[FIND] <food_food>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <wine_glass>', '[GRAB] <wine_glass>', '[DRINK] <wine_glass>']\n}\n", "input": "\nScene_Objects:['blow_dryer', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'remote_control', 'thread', 'computer', 'alcohol', 'bookmark', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'tape', 'standingmirror', 'bed', 'television', 'curtain', 'doorjamb', 'instrument_guitar', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Changing channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[DROP] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'chair', 'walllamp', 'cat', 'wallshelf', 'floor', 'character', 'desk', 'tea_bag', 'toaster', 'plate', 'couch', 'dining_room', 'box', 'newspaper', 'bed', 'toilet_paper', 'curtain', 'coffee_pot', 'mouse', 'kettle', 'cupboard', 'controller', 'clothes_dress', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find Cat in living room.  Walk to cat.  Pet cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['freezer', 'bathroom', 'detergent', 'orchid', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'purse', 'computer', 'table_cloth', 'bookmark', 'stovefan', 'stereo', 'oil', 'closetdrawer', 'coffe_maker', 'television', 'food_donut', 'food_rice', 'tray', 'centerpiece', 'doorjamb', 'pencil', 'maindoor', 'sink', 'home_office', 'table', 'clothes_skirt']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to living room. Find book. Grab book. Open book. Read book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'trashcan', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'thread', 'computer', 'piano_bench', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'oven', 'bed', 'novel', 'check', 'television', 'curtain', 'towel_rack', 'dresser', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Put child in bed, sit down,read book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sponge', 'toilet', 'jelly', 'freezer', 'walllamp', 'bathroom', 'hanger', 'folder', 'bedroom', 'desk', 'tvstand', 'window', 'food_jam', 'microphone', 'dining_room', 'wall_clock', 'powersocket', 'bench', 'shelf', 'bed', 'video_game_controller', 'food_donut', 'water', 'doorjamb', 'cards', 'towel_rack', 'mouse', 'sink', 'colander', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the water, wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <water>", "[FIND] <water>", "[TURNTO] <water>", "[LOOKAT] <water>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tea', 'chair', 'detergent', 'toothbrush', 'orchid', 'knifeblock', 'cpuscreen', 'kitchen_cabinet', 'floor', 'phone', 'bedroom', 'keyboard', 'ceilinglamp', 'faucet', 'window', 'dish_soap', 'food_jam', 'food_orange', 'pillow', 'wall', 'oven', 'bed', 'novel', 'coffe_maker', 'kitchen_counter', 'dishrack', 'cupboard', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTIN] <dish_soap> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'detergent', 'toothbrush', 'drawing', 'cpuscreen', 'food_egg', 'phone', 'character', 'computer', 'piano_bench', 'shoe-shine_kit', 'faucet', 'food_noodles', 'mousepad', 'clothes_hat', 'towel', 'broom', 'cutting_board', 'shelf', 'keys', 'novel', 'food_food', 'check', 'wooden_spoon', 'towel_rack', 'sink', 'table', 'light', 'coffee']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_ice_cream', 'cd_player', 'orchid', 'bedroom', 'purse', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'stereo', 'tvstand', 'balanceball', 'clothes_socks', 'couch', 'food_jam', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'novel', 'check', 'television', 'wine', 'food_onion', 'pencil', 'vacuum_cleaner', 'door', 'sink', 'instrument_violin']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk in the living room, then I go the the stereo and the cd player and turn them on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'freezer', 'nail_polish', 'food_carrot', 'orchid', 'bedroom', 'computer', 'food_bacon', 'toaster', 'plate', 'window', 'clothes_socks', 'closetdrawer', 'oil', 'nightstand', 'basket_for_clothes', 'photoframe', 'check', 'doorjamb', 'towel_rack', 'console', 'dresser', 'door', 'sink', 'table', 'clothes_skirt', 'soap', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[WALK] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'microwave', 'bathroom', 'trashcan', 'toothbrush', 'drawing', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'purse', 'food_bacon', 'ceilinglamp', 'plate', 'laptop', 'deck_of_cards', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'spoon', 'home_office', 'table', 'teeth']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'bathroom', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'shampoo', 'tvstand', 'faucet', 'stereo', 'clothes_socks', 'couch', 'clothes_pants', 'cutting_board', 'basket_for_clothes', 'photoframe', 'food_food', 'video_game_controller', 'coffe_maker', 'coffee_filter', 'television', 'tray', 'curtain', 'shaving_cream', 'laundry_detergent', 'maindoor', 'spoon', 'home_office', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'scissors', 'juice', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'faucet', 'towel', 'dining_room', 'wall', 'photoframe', 'food_food', 'razor', 'television', 'food_rice', 'wooden_spoon', 'after_shave', 'headset', 'laundry_detergent', 'conditioner', 'sink', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Turn on the water faucet. Wet my hands and turn off the faucet. Put soap on my hands. Rub hands together, making sure the soap covers all areas of the hands. Scrub hands for about 20 seconds. Turn on the water water faucet. Rinse hands under the running water. Turn off the water faucet. Dry hands with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'food_pizza', 'juice', 'board_game', 'dog', 'food_butter', 'orchid', 'drawing', 'bathroom_counter', 'bookshelf', 'stovefan', 'mat', 'laptop', 'tvstand', 'picture', 'window', 'couch', 'nightstand', 'wall_clock', 'toy', 'clothes_gloves', 'toilet_paper', 'television', 'ice', 'bathroom_cabinet', 'table', 'light', 'vase', 'instrument_violin']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <home_office>", "[DROP] <plate>", "[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <home_office>", "[DROP] <cup>", "[WALK] <dining_room>", "[WALK] <fork>", "[FIND] <fork>", "[GRAB] <fork>", "[WALK] <home_office>", "[DROP] <fork>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'food_salt', 'microwave', 'freezer', 'orchid', 'drawing', 'hanger', 'button', 'table_cloth', 'tvstand', 'homework', 'couch', 'dining_room', 'nightstand', 'keys', 'coffe_maker', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'diary', 'cards', 'kitchen_counter', 'comb', 'dresser', 'food_steak', 'home_office', 'light', 'clothes_skirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I go to the living room then go to the cat and touch it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'ground_coffee', 'walllamp', 'blow_dryer', 'dough', 'floor', 'ceiling', 'bookshelf', 'toaster', 'faucet', 'wall', 'wall_clock', 'pasta', 'bench', 'powersocket', 'cutting_board', 'box', 'bed', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'sink', 'table', 'clothes_skirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk into the kitchen. I find the coffee maker and open it. I find the coffee filter, grab it and put it into the coffee maker. I find ground coffee, grab it and put it into the coffee maker. I find some water, then I pour it into the coffee maker. I close the coffee maker, I plug it in and I switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'coin', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'toilet', 'chair', 'walllamp', 'bathroom', 'orchid', 'floor', 'ceiling', 'bookshelf', 'table_cloth', 'bookmark', 'ceilinglamp', 'mat', 'stereo', 'shower', 'floor_lamp', 'food_noodles', 'clothes_hat', 'dining_room', 'wall', 'television', 'food_donut', 'doorjamb', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'bathroom_cabinet', 'light', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to wall, find light switch, turn on lights", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'brush', 'mat', 'tvstand', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'bag', 'light', 'dishwasher', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'bowl', 'wall', 'powersocket', 'cup', 'novel', 'coffe_maker', 'food_donut', 'doorjamb', 'mouse', 'dresser', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find a desk and grab a book. Walk to the chair, sit in the chair and read a book.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'dog', 'bathroom', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'computer', 'ceilinglamp', 'window', 'mousepad', 'couch', 'towel', 'wall', 'cup', 'television', 'curtain', 'doorjamb', 'dresser', 'cupboard', 'door', 'sink', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room and turn on the light. I walk to the couch then sit on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'tablelamp', 'blender', 'beer', 'wallshelf', 'floor', 'phone', 'stamp', 'desk', 'stovefan', 'mat', 'laptop', 'stereo', 'window', 'food_noodles', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'cup', 'photoframe', 'clothes_gloves', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'controller', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'detergent', 'trashcan', 'food_carrot', 'knifeblock', 'drawing', 'cpuscreen', 'knife', 'bathroom_counter', 'desk', 'keyboard', 'bookshelf', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'mousepad', 'pillow', 'dining_room', 'oven', 'photoframe', 'food_food', 'kitchen_counter', 'dresser', 'sink', 'home_office']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Walk to the lamp.  Find the switch.  Click the switch off.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[SWITCHOFF] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['scissors', 'toilet', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'envelope', 'computer', 'button', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'standingmirror', 'clothes_gloves', 'television', 'curtain', 'doorjamb', 'towel_rack', 'coffee_pot', 'table', 'soap']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'freezer', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'mop_bucket', 'shower', 'homework', 'candle', 'clothes_pants', 'pillow', 'wall', 'oven', 'bed', 'keys', 'check', 'coffe_maker', 'tray', 'centerpiece', 'pajamas', 'bag', 'spoon', 'soap', 'laser_pointer', 'scrabble']\nTask:Put away keys\nresponse:\n{\n    \"thoughts\": Return keys into a bag,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <keys>', '[FIND] <keys>', '[GRAB] <keys>', '[FIND] <bag>', '[OPEN] <bag>', '[PUTBACK] <keys> <bag>', '[CLOSE] <bag>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'microwave', 'toilet', 'chair', 'walllamp', 'drawing', 'floor', 'ceiling', 'shampoo', 'computer', 'mat', 'food_orange', 'bowl', 'dining_room', 'wall', 'powersocket', 'oven', 'toy', 'shelf', 'cup', 'light_bulb', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'maindoor', 'bathtub', 'light', 'vase']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the kitchen, I find the light switch, and then i switch on the lights in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'beer', 'measuring_cup', 'toilet', 'freezer', 'food_carrot', 'drawing', 'food_egg', 'hanger', 'floor', 'ceiling', 'bedroom', 'mat', 'laptop', 'shower', 'tvstand', 'closetdrawer', 'food_orange', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'drinking_glass', 'wine', 'doorjamb', 'door', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Robot Drinks the water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>', '[PUTOBJBACK] <drinking_glass>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['glue', 'fork', 'tablelamp', 'tea', 'chair', 'freezer', 'nail_polish', 'detergent', 'milk', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'character', 'toaster', 'ceilinglamp', 'stovefan', 'iron', 'window', 'wall', 'powersocket', 'oven', 'standingmirror', 'bed', 'food_donut', 'curtain', 'doorjamb', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <bathroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'wallshelf', 'phone', 'desk', 'toaster', 'stovefan', 'mat', 'shower', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'powersocket', 'bench', 'bed', 'photoframe', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'bathtub', 'sink', 'bathroom_cabinet', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the cat. Pet the cat with your hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'drawing', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'shower', 'electric_shaver', 'food_kiwi', 'wall', 'bench', 'food_cheese', 'photoframe', 'newspaper', 'bed', 'food_food', 'standingmirror', 'television', 'curtain', 'doorjamb', 'shaving_cream', 'door', 'soap', 'clothes_shirt', 'scrabble', 'coffee']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control", "actions": ["[WALK] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'toilet', 'chair', 'blow_dryer', 'food_carrot', 'orchid', 'knifeblock', 'cpuscreen', 'desk', 'clothes_jacket', 'toaster', 'deck_of_cards', 'tvstand', 'dining_room', 'bench', 'basket_for_clothes', 'novel', 'television', 'after_shave', 'coffee_pot', 'mouse', 'cupboard', 'clothes_dress', 'hairdryer', 'home_office', 'colander', 'clothes_shirt', 'washing_machine', 'coffee', 'shoe_rack']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the washing machine. Open the washing machine. Find the basket for clothes. Find the clothes_dress. Grab the clothes_dress. Put the clothes_dress into the washing machine. Find the clothes_shirt. Grab the clothes_shirt. Put the clothes_shirt into the washing machine. Close the washing machine. Switch on the washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <basket_for_clothes>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'chair', 'dog', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'wall', 'powersocket', 'shelf', 'cup', 'doorjamb', 'console', 'mouse', 'shoes', 'dresser', 'bathtub', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I open the room door and press the light button on", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'cleaning_solution', 'faucet', 'balanceball', 'window', 'clothes_socks', 'food_orange', 'wall', 'powersocket', 'bench', 'cutting_board', 'photoframe', 'novel', 'check', 'rag', 'lighter', 'mouse', 'kitchen_counter', 'hair', 'sink', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'scrabble', 'dirt']\nTask:Wash sink\nresponse:\n{\n    \"thoughts\": Turn on faucet. Rub down sink.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <sink>', '[POUR] <cleaning_solution> <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[SCRUB] <sink>', '[WASH] <sink>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'needle', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'stamp', 'envelope', 'computer', 'mat', 'ceilinglamp', 'oil', 'food_orange', 'wall', 'bench', 'shelf', 'food_cheese', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to room. I enter room. I find where the lightswitch is. I turn on the lights with the lightswitch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[LOOKAT] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'crayon', 'walllamp', 'trashcan', 'dough', 'floor', 'phone', 'bedroom', 'tea_bag', 'toaster', 'mat', 'plate', 'ceilinglamp', 'dish_soap', 'wall', 'pasta', 'oven', 'cutting_board', 'food_cheese', 'newspaper', 'coffe_maker', 'check', 'food_food', 'tray', 'towel_rack', 'pencil', 'dresser', 'conditioner', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['fork', 'dry_pasta', 'filing_cabinet', 'scissors', 'kitchen_cabinet', 'envelope', 'faucet', 'dish_soap', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall_clock', 'bench', 'toy', 'cutting_board', 'razor', 'food_onion', 'doorjamb', 'towel_rack', 'kitchen_counter', 'pajamas', 'clothes_dress', 'bathtub', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'vase', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTIN] <dish_soap> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'tablelamp', 'blender', 'tea', 'chair', 'freezer', 'drawing', 'orchid', 'hanger', 'floor', 'shampoo', 'stamp', 'bedroom', 'desk', 'computer', 'mat', 'stereo', 'mousepad', 'food_jam', 'dining_room', 'wall', 'wall_clock', 'bench', 'photoframe', 'food_food', 'dresser', 'napkin', 'home_office', 'table', 'spoon']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'food_carrot', 'drawing', 'floor', 'thread', 'toaster', 'mat', 'picture', 'oil', 'towel', 'nightstand', 'powersocket', 'toy', 'shelf', 'box', 'cup', 'cutting_board', 'bed', 'photoframe', 'novel', 'toilet_paper', 'check', 'television', 'tray', 'food_onion', 'curtain', 'dresser', 'door', 'hairdryer', 'tooth_paste']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the radio; switch on the radio and listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'shoes', 'door', 'table', 'light', 'creditcard']\nTask:Watch  TV\nresponse:\n{\n    \"thoughts\": Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'bathroom', 'food_carrot', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'shampoo', 'character', 'stamp', 'keyboard', 'mat', 'faucet', 'picture', 'food_peanut_butter', 'dining_room', 'pasta', 'bench', 'cutting_board', 'cup', 'standingmirror', 'photoframe', 'coin', 'cards', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'vase']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'nail_polish', 'knifeblock', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'desk', 'cleaning_solution', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'shower', 'faucet', 'ceilinglamp', 'pillow', 'bench', 'cup', 'mouse', 'controller', 'laundry_detergent', 'door', 'napkin', 'light', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_bread', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'stovefan', 'mat', 'tvstand', 'window', 'closetdrawer', 'microphone', 'food_orange', 'pillow', 'wall', 'nightstand', 'bed', 'food_rice', 'curtain', 'kettle', 'dresser', 'pajamas', 'bathtub', 'bathroom_cabinet', 'table', 'vase']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <sheets>", "[PULL] <sheets>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'toaster', 'window', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'wall_clock', 'television', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'clothes_shirt']\nTask:Organize closet\nresponse:\n{\n    \"thoughts\": Organize the hanging clothes in my closet by color and type and get rid of clothes I don't wear.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[CLOSE] <dresser>', '[WALK] <basket_for_clothes>', '[PUTBACK] <clothes_shirt> <basket_for_clothes>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'tablelamp', 'mop', 'microwave', 'freezer', 'walllamp', 'needle', 'food_egg', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'stovefan', 'stereo', 'form', 'dining_room', 'wall', 'oven', 'powersocket', 'cutting_board', 'toy', 'keys', 'food_food', 'coffe_maker', 'wine', 'tray', 'kitchen_counter', 'clothes_scarf', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'tablelamp', 'ground_coffee', 'walllamp', 'trashcan', 'knifeblock', 'longboard', 'floor', 'ceiling', 'character', 'bookmark', 'toaster', 'tvstand', 'wall', 'clothes_gloves', 'keys', 'coffe_maker', 'check', 'video_game_controller', 'coffee_filter', 'toilet_paper', 'food_onion', 'curtain', 'kitchen_counter', 'controller', 'hairdryer', 'dishwasher', 'vase', 'soap']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": You walk into the kitchen and turn to the coffee maker. You grab the coffee filter and put into the coffee machine, you also grab the ground coffee and put it into the coffee machine. You turn on the coffee machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'drawing', 'floor', 'character', 'clothes_jacket', 'mop_bucket', 'mat', 'ceilinglamp', 'window', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'oven', 'cutting_board', 'novel', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'shoes', 'hairbrush', 'dresser', 'maindoor', 'sink', 'napkin', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Sitting down and relaxing on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'mop', 'microwave', 'ground_coffee', 'walllamp', 'cat', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'toaster', 'brush', 'tvstand', 'shower', 'faucet', 'food_jam', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'mouthwash', 'coffee_filter', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'dishwasher', 'scrabble']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Make coffee in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[PLUGIN] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'measuring_cup', 'dog', 'walllamp', 'bathroom', 'floor', 'phone', 'bedroom', 'mat', 'stereo', 'faucet', 'window', 'closetdrawer', 'dining_room', 'food_cheese', 'standingmirror', 'bed', 'check', 'television', 'coffee_filter', 'food_rice', 'ice', 'curtain', 'doorjamb', 'pencil', 'conditioner', 'bathtub', 'napkin', 'vase']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'bookshelf', 'brush', 'mat', 'stereo', 'picture', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'shelf', 'cup', 'television', 'wine', 'curtain', 'food_onion', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'home_office', 'table']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to living room. Walk to radio. Switch on the radio. Walk to sofa and sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'computer', 'stovefan', 'mat', 'faucet', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'newspaper', 'photoframe', 'television', 'curtain', 'rag', 'doorjamb', 'mouse', 'pencil', 'table', 'vase', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I go to the living room and sit in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'toilet', 'jelly', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'stovefan', 'mat', 'shower', 'food_kiwi', 'wall', 'nightstand', 'oven', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'water', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'door', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_pizza', 'filing_cabinet', 'tablelamp', 'freezer', 'nail_polish', 'cpuscreen', 'desk', 'bookshelf', 'keyboard', 'alcohol', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'wall', 'wall_clock', 'bed', 'coffe_maker', 'mouthwash', 'video_game_controller', 'television', 'razor', 'check', 'tray', 'curtain', 'cd', 'spoon', 'vase']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find the cat. Walk over to the cat. Touch the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'console', 'dresser', 'cupboard', 'home_office', 'table', 'clothes_skirt', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'foundation', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'napkin', 'light', 'table', 'dirt']\nTask:Work\nresponse:\n", "output": {"thoughts": "go to home office. turn the computer on. begin working", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'food_egg', 'clothes_underwear', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'bench', 'box', 'television', 'curtain', 'towel_rack', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Changing channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[DROP] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'toilet', 'chair', 'freezer', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'character', 'bookshelf', 'mat', 'window', 'couch', 'dustpan', 'pillow', 'towel', 'wall', 'bench', 'food_food', 'television', 'coin', 'tray', 'curtain', 'doorjamb', 'bathtub', 'spoon', 'table', 'soap']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music", "actions": ["[WALK] <home_office>", "[WALK] <dvd_player>", "[FIND] <dvd_player>", "[TURNTO] <dvd_player>", "[LOOKAT] <dvd_player>", "[SWITCHON] <dvd_player>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'chair', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'food_cake', 'envelope', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'wall_clock', 'light_bulb', 'television', 'curtain', 'hair', 'food_steak', 'sink', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I enter the living room and sit on the couch. I grab the remote and turn on the TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'ground_coffee', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'powersocket', 'check', 'television', 'doorjamb', 'mouse', 'pencil', 'dresser', 'sink', 'light', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "I go to the home office to work on my project. I turn on the light and walk to my desk. I turn on the computer. Then I look through the documents. After I'm done, I turn off the computer.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <check>", "[TURNTO] <check>", "[LOOKAT] <check>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'toilet', 'dog', 'detergent', 'orchid', 'wallshelf', 'food_egg', 'cpuscreen', 'phone', 'stamp', 'bedroom', 'bookshelf', 'bookmark', 'shower', 'mousepad', 'food_jam', 'towel', 'pasta', 'bed', 'novel', 'food_food', 'curtain', 'centerpiece', 'shaving_cream', 'dresser', 'cupboard', 'table', 'creditcard']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the entrance hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['tea', 'microwave', 'toilet', 'jelly', 'bathroom_counter', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'food_noodles', 'clothes_socks', 'mousepad', 'closetdrawer', 'food_kiwi', 'wall', 'cutting_board', 'cup', 'coffe_maker', 'toothbrush_holder', 'television', 'oven_mitts', 'water', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'soap', 'tooth_paste']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, walk to cupboard, open cupboard, grab cup, walk to sink, pour water in cup, drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'bedroom', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'check', 'television', 'curtain', 'doorjamb', 'dresser', 'door', 'table', 'colander']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'food_salt', 'toilet', 'walllamp', 'bathroom', 'food_butter', 'drawing', 'hanger', 'shampoo', 'bedroom', 'purse', 'computer', 'stovefan', 'laptop', 'ceilinglamp', 'electric_shaver', 'window', 'homework', 'clothes_socks', 'wall_clock', 'shelf', 'basket_for_clothes', 'bed', 'coffe_maker', 'check', 'wooden_spoon', 'centerpiece', 'lighter', 'kettle']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "The task is browse and surf on the internet regarding a particular topic", "actions": ["[WALK] <bedroom>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[WATCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'food_salt', 'beer', 'microwave', 'trashcan', 'food_carrot', 'milk', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'ceilinglamp', 'tvstand', 'balanceball', 'deck_of_cards', 'window', 'mousepad', 'couch', 'wall', 'bench', 'television', 'curtain', 'mouse', 'controller', 'clothes_scarf', 'napkin', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit on the couch and watch TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'dog', 'nail_polish', 'orchid', 'drawing', 'floor', 'ceiling', 'character', 'desk', 'computer', 'clothes_jacket', 'stovefan', 'mat', 'tvstand', 'oil', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'newspaper', 'food_food', 'doorjamb', 'shoes', 'maindoor', 'bathtub', 'light', 'tooth_paste', 'clothes_shirt', 'scrabble']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the dining room and turn on the lights", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'detergent', 'drawing', 'knifeblock', 'food_egg', 'floor', 'computer', 'toaster', 'stovefan', 'tvstand', 'stereo', 'window', 'food_noodles', 'mousepad', 'couch', 'pillow', 'dining_room', 'nightstand', 'pasta', 'bed', 'oven_mitts', 'tray', 'curtain', 'lighter', 'towel_rack', 'shoes', 'cd', 'pajamas']\nTask:Put clothes away\nresponse:\n{\n    \"thoughts\": I take clean clothes and either place them in dresser drawers or hang them in a closet.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'tea', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'mop_bucket', 'stovefan', 'mat', 'tvstand', 'faucet', 'window', 'clothes_pants', 'wall', 'wall_clock', 'oven', 'toy', 'cup', 'water', 'colander', 'tray', 'centerpiece', 'cd', 'maindoor', 'sink', 'bathroom_cabinet', 'home_office', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Walk to sink and turn on the water. Grab the soap and wash the hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <water>", "[TURNTO] <water>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'ground_coffee', 'food_butter', 'toothbrush', 'knifeblock', 'drawing', 'phone', 'keyboard', 'bookshelf', 'table_cloth', 'mat', 'plate', 'electric_shaver', 'window', 'pillow', 'nightstand', 'oven', 'newspaper', 'clothes_gloves', 'photoframe', 'toilet_paper', 'water', 'centerpiece', 'towel_rack', 'kettle', 'pencil', 'clothes_scarf', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[FIND] <toothbrush>', '[FIND] <water>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'bookmark', 'toaster', 'laptop', 'stovefan', 'electric_shaver', 'food_peanut_butter', 'dining_room', 'nightstand', 'oven', 'standingmirror', 'oven_mitts', 'food_onion', 'doorjamb', 'paper_towel', 'shoes', 'maindoor', 'vase', 'hands_both']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walking towards Toilet room, And dispose of human waste. Cleaning your hands with tissue paper or water.", "actions": ["[WALK] <bathroom>", "[WALK] <paper_towel>", "[FIND] <paper_towel>", "[TURNTO] <paper_towel>", "[LOOKAT] <paper_towel>", "[GRAB] <paper_towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'microwave', 'chair', 'band-aids', 'woman', 'needle', 'longboard', 'man', 'keyboard', 'envelope', 'toaster', 'mat', 'ceilinglamp', 'shower', 'faucet', 'stereo', 'nightstand', 'box', 'light_bulb', 'coffe_maker', 'curtain', 'towel_rack', 'shoes', 'dresser', 'hairbrush', 'door', 'vase', 'instrument_violin', 'creditcard']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'band-aids', 'freezer', 'bathroom', 'orchid', 'needle', 'wallshelf', 'knife', 'hanger', 'floor', 'ceiling', 'bedroom', 'keyboard', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'wall', 'oven', 'food_food', 'oven_mitts', 'doorjamb', 'pencil', 'cd', 'dresser', 'door', 'maindoor', 'bathroom_cabinet', 'light', 'clothes_skirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the bedroom and use the button on the wall, next to the door to turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'mop', 'microwave', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'newspaper', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'measuring_cup', 'chair', 'nail_polish', 'cat', 'drawing', 'hanger', 'floor', 'phone', 'desk', 'keyboard', 'food_bacon', 'table_cloth', 'bookmark', 'stovefan', 'plate', 'shower', 'dish_soap', 'couch', 'wall', 'powersocket', 'basket_for_clothes', 'toilet_paper', 'pencil', 'hair', 'home_office', 'light', 'dishwasher', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'towel', 'wall', 'wall_clock', 'oven', 'television', 'tray', 'rag', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <bathroom_cabinet>', '[FIND] <shower>', '[RINSE] <shower>']\n}\n", "input": "\nScene_Objects:['pot', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'computer', 'button', 'brush', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'standingmirror', 'coffe_maker', 'television', 'food_rice', 'wine', 'curtain', 'doorjamb', 'diary', 'pencil', 'door', 'bathroom_cabinet', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'freezer', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'microphone', 'pillow', 'wall', 'wall_clock', 'nightstand', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'home_office', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room and sit on the couch, ease into the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'dog', 'orchid', 'floor', 'desk', 'button', 'clothes_jacket', 'mat', 'stereo', 'shower', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toilet_paper', 'mouthwash', 'television', 'curtain', 'after_shave', 'doorjamb', 'hair', 'food_steak', 'napkin', 'home_office', 'table', 'vase']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "sitting down to listen to music on the radio", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'trashcan', 'ceilingfan', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'laptop', 'balanceball', 'mousepad', 'pillow', 'wall', 'cup', 'bed', 'television', 'razor', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'spoon', 'light', 'vase']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": push any ligth button to turn it on, for example in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'beer', 'chair', 'bills', 'knife', 'hanger', 'bathroom_counter', 'floor', 'computer', 'stovefan', 'mat', 'tvstand', 'window', 'couch', 'pillow', 'towel', 'dining_room', 'food_kiwi', 'wall', 'toy', 'photoframe', 'novel', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'bag', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on couch. Grab book. Open book. Read book.", "actions": ["[WALK] <couch>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'bathroom', 'ceilingfan', 'knife', 'drawing', 'needle', 'knifeblock', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'phone', 'piano_bench', 'faucet', 'window', 'teeth', 'clothes_pants', 'wall', 'photoframe', 'check', 'oven_mitts', 'tray', 'pencil', 'clothes_dress', 'sink', 'bathroom_cabinet', 'home_office', 'clothes_skirt', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[PUSH] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'freezer', 'toothbrush', 'hanger', 'floor', 'ceiling', 'bedroom', 'table_cloth', 'stovefan', 'toaster', 'tvstand', 'mousepad', 'couch', 'pillow', 'nightstand', 'toothbrush_holder', 'tray', 'towel_rack', 'kettle', 'vacuum_cleaner', 'conditioner', 'door', 'bathtub', 'sink', 'food_steak', 'maindoor', 'soap', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, take toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now brush teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[TURNTO] <toothbrush>", "[LOOKAT] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'drawing', 'hanger', 'floor', 'phone', 'food_bacon', 'stovefan', 'plate', 'tvstand', 'faucet', 'couch', 'food_jam', 'wall', 'bench', 'powersocket', 'photoframe', 'food_food', 'oven_mitts', 'doorjamb', 'mouse', 'kitchen_counter', 'headset', 'cd', 'dresser', 'cupboard', 'clothes_dress', 'door', 'table', 'light']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": Sit down at the dinning table with my family and eat dinner.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[GRAB] <food_food>', '[FIND] <plate>', '[PUTBACK] <food_food> <plate>', '[GRAB] <food_food>', '[PUTOBJBACK] <food_food>', '[FIND] <fork>', '[GRAB] <fork>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'filing_cabinet', 'sheets', 'detergent', 'milk', 'hanger', 'phone', 'desk', 'clothes_jacket', 'plate', 'dish_soap', 'mousepad', 'couch', 'candle', 'pillow', 'bowl', 'basket_for_clothes', 'standingmirror', 'keys', 'food_food', 'check', 'razor', 'coffee_filter', 'mouse', 'kitchen_counter', 'controller', 'door', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Put dishes in the dishwasher and start the dishwasher.", "actions": ["[RUN] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[WALK] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'milk', 'wallshelf', 'cpuscreen', 'hanger', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'table_cloth', 'stovefan', 'tvstand', 'window', 'closetdrawer', 'candle', 'dining_room', 'wall_clock', 'oven', 'bench', 'photoframe', 'television', 'tray', 'kitchen_counter', 'door', 'bathtub', 'sink', 'table', 'soap', 'creditcard']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'food_pizza', 'blow_dryer', 'bathroom', 'dough', 'computer', 'ceilinglamp', 'stereo', 'foundation', 'couch', 'pasta', 'powersocket', 'toy', 'cutting_board', 'food_food', 'check', 'coffe_maker', 'after_shave', 'rag', 'curtain', 'shaving_cream', 'shoes', 'controller', 'door', 'bathtub', 'sink', 'bathroom_cabinet', 'laser_pointer', 'creditcard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch, flip the switch to the on position", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'detergent', 'cat', 'milk', 'phone', 'ceiling', 'desk', 'mop_bucket', 'brush', 'ceilinglamp', 'tvstand', 'faucet', 'toaster', 'mat', 'closetdrawer', 'microphone', 'nightstand', 'wall_clock', 'toy', 'cup', 'food_food', 'after_shave', 'lighter', 'towel_rack', 'mouse', 'pajamas', 'controller', 'laundry_detergent', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Go to the living room where the cat sleeps. Find cat. Proceed to petting cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['juice', 'blow_dryer', 'cat', 'toothbrush', 'orchid', 'floor', 'keyboard', 'mat', 'ceilinglamp', 'shoe-shine_kit', 'picture', 'window', 'mousepad', 'couch', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'shelf', 'photoframe', 'television', 'ice', 'curtain', 'after_shave', 'doorjamb', 'diary', 'vacuum_cleaner', 'conditioner', 'home_office', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I would walk into the living room. Then sit on the couch and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'foundation', 'dustpan', 'pillow', 'broom', 'wall', 'powersocket', 'cup', 'video_game_controller', 'television', 'doorjamb', 'hairbrush', 'door', 'sink', 'bag', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toothbrush', 'knife', 'drawing', 'desk', 'bookshelf', 'purse', 'computer', 'table_cloth', 'brush', 'tvstand', 'shoe-shine_kit', 'picture', 'faucet', 'wall', 'bench', 'cutting_board', 'shelf', 'light_bulb', 'razor', 'after_shave', 'doorjamb', 'diary', 'kitchen_counter', 'dishrack', 'dresser', 'home_office', 'table', 'vase', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the sink. Take dishes out of sink. Open dishwasher and place dishes in dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'cd_player', 'toilet', 'blow_dryer', 'floor', 'desk', 'tea_bag', 'remote_control', 'button', 'toaster', 'plate', 'tvstand', 'couch', 'nightstand', 'cup', 'photoframe', 'keys', 'novel', 'check', 'diary', 'vacuum_cleaner', 'cd', 'comb', 'hair', 'clothes_scarf', 'home_office', 'light', 'table', 'creditcard']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[PLUGIN] <cd_player>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'food_butter', 'floor', 'food_cake', 'thread', 'mat', 'shower', 'faucet', 'window', 'oil', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'bench', 'television', 'curtain', 'doorjamb', 'lighter', 'towel_rack', 'dresser', 'hairbrush', 'laundry_detergent', 'door', 'bathroom_cabinet', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I get home from work and walk into the living room. I see the couch and go to sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'jelly', 'bathroom', 'knife', 'dvd_player', 'character', 'bedroom', 'thread', 'computer', 'toaster', 'shoe-shine_kit', 'deck_of_cards', 'shower', 'electric_shaver', 'homework', 'food_peanut_butter', 'form', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'novel', 'light_bulb', 'food_food', 'curtain', 'coffee_pot', 'food_steak', 'door', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the radio. Turn on the radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'ground_coffee', 'food_butter', 'dough', 'drawing', 'floor', 'desk', 'bookmark', 'mat', 'faucet', 'floor_lamp', 'window', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'novel', 'television', 'curtain', 'doorjamb', 'chessboard', 'controller', 'maindoor', 'home_office', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "First I walk into the living room, I find the lamp and switch it on. Then I find a book, grab it, sit in the sofa, open the book and read it. When I'm finished, I find a bookmark, grab it and put it in the book. Then I close the book, put it back, and switch off the lamp.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[PUTBACK] <bookmark> <novel>", "[PUTOBJBACK] <novel>", "[SWITCHOFF] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'mop', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office']\nTask:Social media  checks\nresponse:\n{\n    \"thoughts\": Logging on to  things like  facebook  and  twitter,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['juice', 'toilet', 'chair', 'bathroom', 'drawing', 'knife', 'wallshelf', 'floor', 'ceiling', 'stamp', 'bookshelf', 'envelope', 'computer', 'shower', 'window', 'wall', 'nightstand', 'oven', 'toy', 'shelf', 'cup', 'standingmirror', 'coffe_maker', 'toothbrush_holder', 'food_donut', 'after_shave', 'dresser', 'bathtub', 'home_office', 'light']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to bathroom, find toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'ground_coffee', 'knifeblock', 'hanger', 'floor', 'ceiling', 'purse', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'wall', 'cup', 'razor', 'coin', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": To go kitchen and find a cup. open the faucet and fill water in the cup and drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <cup>', '[GRAB] <cup>', '[TURNTO] <faucet>', '[POINTAT] <faucet>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'walllamp', 'food_carrot', 'knife', 'wallshelf', 'thread', 'envelope', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'couch', 'pillow', 'broom', 'wall', 'bench', 'powersocket', 'tape', 'shelf', 'basket_for_clothes', 'photoframe', 'bed', 'oven_mitts', 'towel_rack', 'shaving_cream', 'kitchen_counter', 'shoes', 'table', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find a radio there and switch it on to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'dry_pasta', 'walllamp', 'trashcan', 'food_carrot', 'dough', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'wall', 'wall_clock', 'oven', 'food_cheese', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher']\nTask:Keep an eye on stove as something is cooking\nresponse:\n{\n    \"thoughts\": Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <pot>', '[TURNTO] <pot>', '[LOOKAT] <pot>', '[FIND] <wall_clock>', '[TURNTO] <wall_clock>', '[LOOKAT] <wall_clock>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['mop', 'microwave', 'freezer', 'ground_coffee', 'dough', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'bedroom', 'tea_bag', 'toaster', 'faucet', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toy', 'coffe_maker', 'television', 'coffee_filter', 'water', 'mouse', 'kitchen_counter', 'sink', 'table']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'trashcan', 'desk', 'keyboard', 'stovefan', 'ceilinglamp', 'mat', 'toaster', 'stereo', 'window', 'couch', 'dining_room', 'nightstand', 'wall', 'cutting_board', 'bed', 'photoframe', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'hairdryer', 'bathtub', 'bathroom_cabinet', 'home_office', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk into room, grab phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['food_bread', 'food_pizza', 'microwave', 'dog', 'band-aids', 'orchid', 'wallshelf', 'floor', 'desk', 'mat', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'tape', 'cutting_board', 'bed', 'standingmirror', 'television', 'curtain', 'doorjamb', 'conditioner', 'sink', 'home_office', 'table', 'light', 'bathroom_cabinet', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'clothes_jacket', 'table_cloth', 'mat', 'floor_lamp', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'light_bulb', 'novel', 'mouthwash', 'television', 'ice', 'curtain', 'after_shave', 'doorjamb', 'console', 'kettle', 'conditioner', 'bathtub', 'sink', 'table', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['mop', 'toilet', 'chair', 'walllamp', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'envelope', 'computer', 'clothes_jacket', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'pasta', 'powersocket', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'home_office', 'light', 'table', 'vase']\nTask:Work\nresponse:\n", "output": {"thoughts": "Type on a computer in the home office.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'purse', 'toaster', 'faucet', 'shower', 'homework', 'pillow', 'wall', 'oven', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to kitchen, turn on water, find soap, apply to hand, lather, rinse, dry hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'food_butter', 'floor', 'ceiling', 'desk', 'purse', 'mop_bucket', 'stovefan', 'deck_of_cards', 'iron', 'shower', 'stereo', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'cutting_board', 'bed', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'door', 'hairdryer', 'light', 'table', 'laser_pointer', 'scrabble']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go into my room, find the light,  then turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'beer', 'freezer', 'trashcan', 'orchid', 'floor', 'ceiling', 'bedroom', 'computer', 'toaster', 'window', 'couch', 'candle', 'dining_room', 'broom', 'oven', 'toy', 'check', 'video_game_controller', 'tray', 'kettle', 'coffee_pot', 'mouse', 'cd', 'cupboard', 'controller', 'bathroom_cabinet', 'light', 'dishwasher', 'laser_pointer']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I pick up my phone. I touch it to navigate to the web browser. I can swipe and tap the screen to visit different websites. Then I read or watch the content within those websites.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[TOUCH] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'walllamp', 'bathroom', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'stovefan', 'ceilinglamp', 'plate', 'shower', 'tvstand', 'food_noodles', 'bowl', 'wall', 'tape', 'shelf', 'bed', 'check', 'doorjamb', 'kitchen_counter', 'shoes', 'sink', 'light', 'laser_pointer']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "walk to bathroom , find shower , take a shower", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'trashcan', 'bills', 'knife', 'floor', 'desk', 'mail', 'button', 'toaster', 'plate', 'faucet', 'shower', 'foundation', 'couch', 'bowl', 'wall', 'nightstand', 'oven', 'bed', 'clothes_gloves', 'novel', 'coffe_maker', 'check', 'tray', 'sink', 'table', 'dishwasher', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['blender', 'toilet', 'chair', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'window', 'mousepad', 'broom', 'wall', 'wall_clock', 'nightstand', 'novel', 'coin', 'lighter', 'centerpiece', 'mouse', 'dresser']\nTask:Read book\nresponse:\n", "output": {"thoughts": "find a book to read. sit down. read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sheets', 'filing_cabinet', 'food_salt', 'toilet', 'band-aids', 'walllamp', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'phone', 'computer', 'mop_bucket', 'stovefan', 'tvstand', 'deck_of_cards', 'closetdrawer', 'newspaper', 'coffe_maker', 'doorjamb', 'dresser', 'cupboard', 'door', 'food_steak', 'bag', 'sink', 'table', 'dishwasher', 'bathroom_cabinet']\nTask:Put umbrella away\nresponse:\n{\n    \"thoughts\": Place the umbrella in the umbrella bucket point first.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mop_bucket>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'blow_dryer', 'hanger', 'floor', 'phone', 'desk', 'bookshelf', 'food_cake', 'computer', 'ceilinglamp', 'tvstand', 'shower', 'window', 'dining_room', 'wall', 'bench', 'shelf', 'coffe_maker', 'food_food', 'rag', 'kettle', 'pencil', 'cd', 'home_office', 'light', 'bathroom_cabinet', 'instrument_violin']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. I walk over to the toilet. I sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'water_glass', 'pot', 'walllamp', 'bathroom', 'milk', 'knife', 'knifeblock', 'drawing', 'floor', 'ceiling', 'bedroom', 'toaster', 'stovefan', 'plate', 'shower', 'faucet', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'coffee_cup', 'dishwasher', 'table']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and put plates, cutlery and glasses into the dishwasher, put the soap, close the machine and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <dishwasher>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <dishwasher>', '[FIND] <pot>', '[GRAB] <pot>', '[PUTBACK] <pot> <dishwasher>', '[FIND] <pot>', '[GRAB] <pot>', '[PUTBACK] <pot> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[PUTBACK] <water_glass> <dishwasher>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[PUTBACK] <water_glass> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'detergent', 'floor', 'ceiling', 'toaster', 'faucet', 'stereo', 'electric_shaver', 'window', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'cutting_board', 'coffe_maker', 'television', 'drinking_glass', 'after_shave', 'shaving_cream', 'pencil', 'kitchen_counter', 'comb', 'cupboard', 'sink', 'bathroom_cabinet', 'vase', 'clothes_shirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'chair', 'ground_coffee', 'food_butter', 'floor', 'ceiling', 'computer', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'brush', 'stereo', 'mat', 'electric_shaver', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'tape', 'cup', 'coffe_maker', 'coffee_filter', 'water', 'towel_rack', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'blow_dryer', 'cat', 'wallshelf', 'ceiling', 'desk', 'bedroom', 'keyboard', 'stovefan', 'mousepad', 'powersocket', 'oven', 'bed', 'food_food', 'check', 'wooden_spoon', 'towel_rack', 'mouse', 'dishrack', 'comb', 'maindoor', 'bathtub', 'napkin', 'home_office', 'light', 'table', 'soap', 'laser_pointer', 'scrabble']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <address_book>", "[FIND] <address_book>", "[GRAB] <address_book>", "[READ] <address_book>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_salt', 'chair', 'ground_coffee', 'bathroom', 'floor', 'ceiling', 'button', 'mop_bucket', 'homework', 'window', 'clothes_hat', 'clothes_pants', 'wall', 'toy', 'novel', 'coffe_maker', 'food_food', 'razor', 'wooden_spoon', 'doorjamb', 'diary', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'coffee_cup', 'dishwasher', 'clothes_skirt', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to the kitchen at the coffee maker, grab coffee, pour coffee into coffee maker, put coffee into coffee maker, push coffee maker's button, wait to complete, grab coffee cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[OPEN] <coffe_maker>', '[POUR] <coffee> <coffe_maker>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'bathroom', 'detergent', 'cat', 'food_butter', 'toothbrush', 'orchid', 'cpuscreen', 'desk', 'bedroom', 'purse', 'envelope', 'mail', 'mop_bucket', 'toaster', 'shower', 'faucet', 'oil', 'food_orange', 'powersocket', 'cutting_board', 'shelf', 'curtain', 'after_shave', 'diary', 'kitchen_counter', 'door', 'napkin', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "before going to bed or when wake up earily morning brush teeth", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'walllamp', 'food_butter', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'stereo', 'faucet', 'window', 'mousepad', 'pillow', 'wall', 'powersocket', 'bed', 'coffee_table', 'television', 'doorjamb', 'pencil', 'dresser', 'clothes_dress', 'door', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room and look at the coffee table. Grab the remote control and turn to the television and point the remote control. Put back the remote control.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <coffee_table>', '[TURNTO] <coffee_table>', '[LOOKAT] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'purse', 'stovefan', 'mat', 'toaster', 'faucet', 'shower', 'plate', 'closetdrawer', 'towel', 'wall', 'wall_clock', 'oven', 'toy', 'hands_both', 'coin', 'tray', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom. Turn on the sink. Rinse hands. Apply soap and scrub your hands together. Rinse hands again. Turn off faucet. Dry hands on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'chair', 'bathroom', 'trashcan', 'knifeblock', 'orchid', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'phone', 'desk', 'food_cake', 'computer', 'table_cloth', 'tvstand', 'faucet', 'couch', 'oven', 'powersocket', 'cup', 'photoframe', 'check', 'tray', 'curtain', 'wooden_spoon', 'doorjamb', 'door', 'bathtub', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk into the home office. Pick up the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'jelly', 'freezer', 'walllamp', 'drawing', 'character', 'tea_bag', 'stovefan', 'mat', 'tvstand', 'stereo', 'shoe-shine_kit', 'window', 'ceilinglamp', 'dustpan', 'food_jam', 'pillow', 'wall', 'wall_clock', 'newspaper', 'coffe_maker', 'food_donut', 'doorjamb', 'coffee_pot', 'cards', 'door', 'bathroom_cabinet', 'colander', 'scrabble']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I reach out arm. I put hand on radio. I turn on radio. I pick a station.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[TOUCH] <stereo>", "[TURNTO] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'jelly', 'walllamp', 'bathroom', 'knife', 'wallshelf', 'floor', 'ceiling', 'bookmark', 'ceilinglamp', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'television', 'toothbrush_holder', 'wooden_spoon', 'kettle', 'shoes', 'cupboard', 'conditioner', 'clothes_dress', 'door', 'napkin', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Pick up a book and open to the last page you were on. Read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <bookmark>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'blow_dryer', 'food_carrot', 'drawing', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'toaster', 'shoe-shine_kit', 'window', 'oil', 'food_orange', 'wall', 'wall_clock', 'food_food', 'curtain', 'after_shave', 'doorjamb', 'chessboard', 'door', 'bathtub', 'napkin', 'bathroom_cabinet', 'light', 'vase']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone. .I reach my arm out. I open my hand. I place my hand on the phone. I close my fingers around the phone. I pick up the phone.", "actions": ["[WALK] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'sheets', 'food_ice_cream', 'food_salt', 'measuring_cup', 'microwave', 'toilet', 'detergent', 'ceilingfan', 'drawing', 'hanger', 'floor', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'plate', 'window', 'floor_lamp', 'mousepad', 'couch', 'pillow', 'wall', 'clothes_gloves', 'novel', 'check', 'television', 'food_rice', 'doorjamb', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'chair', 'freezer', 'food_butter', 'drawing', 'bills', 'desk', 'tea_bag', 'keyboard', 'purse', 'food_cake', 'envelope', 'plate', 'dish_soap', 'mousepad', 'pillow', 'wall_clock', 'powersocket', 'shelf', 'standingmirror', 'check', 'coffe_maker', 'doorjamb', 'towel_rack', 'shaving_cream', 'kitchen_counter', 'dresser', 'sink']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Pick up dish.  Put in soapy water.  Wash dish.", "actions": ["[WALK] <dining_room>", "[WALK] <dish_soap>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <dish_soap> <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[WIPE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'crayon', 'chair', 'nail_polish', 'cat', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'bookshelf', 'computer', 'food_bacon', 'clothes_jacket', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'toy', 'doorjamb', 'kettle', 'mouse', 'dresser', 'hairdryer', 'sink', 'bathroom_cabinet', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Coming home turning on lights.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'milk', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'shower', 'window', 'couch', 'clothes_hat', 'pillow', 'wall', 'photoframe', 'check', 'mouthwash', 'razor', 'television', 'coin', 'coffe_maker', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and I lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'longboard', 'character', 'bedroom', 'computer', 'faucet', 'homework', 'clothes_socks', 'food_noodles', 'dining_room', 'photoframe', 'check', 'television', 'after_shave', 'rag', 'lighter', 'towel_rack', 'kitchen_counter', 'dresser', 'clothes_dress', 'home_office', 'light', 'laser_pointer', 'scrabble']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office. I find the telephone. I pick up the telephone handset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['microwave', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'bedroom', 'computer', 'alcohol', 'mat', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'bench', 'cutting_board', 'standingmirror', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'chessboard', 'dresser', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room. Sit on the couch. Turn on TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'freezer', 'walllamp', 'bathroom', 'hanger', 'phone', 'desk', 'bookshelf', 'ceilinglamp', 'shoe-shine_kit', 'balanceball', 'faucet', 'couch', 'nightstand', 'wall', 'bench', 'powersocket', 'clothes_gloves', 'check', 'television', 'food_onion', 'after_shave', 'curtain', 'shaving_cream', 'mouse', 'maindoor', 'sink', 'bag', 'home_office']\nTask:Open front door\nresponse:\n{\n    \"thoughts\": Walk over to the front door, turn the knob, and open the door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'microwave', 'freezer', 'walllamp', 'ground_coffee', 'food_carrot', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'stovefan', 'mat', 'faucet', 'iron', 'oil', 'closetdrawer', 'wall', 'wall_clock', 'oven', 'powersocket', 'cutting_board', 'food_cheese', 'standingmirror', 'coffe_maker', 'food_food', 'wine', 'tray', 'kitchen_counter', 'sink', 'bathroom_cabinet']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "go to the kitchen and cook some thing to eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <oven>", "[OPEN] <oven>", "[FIND] <food_pizza>", "[GRAB] <food_pizza>", "[PUTIN] <food_pizza> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'detergent', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'thread', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'photoframe', 'toilet_paper', 'television', 'razor', 'coin', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Push power button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'food_pizza', 'measuring_cup', 'microwave', 'nail_polish', 'bathroom', 'dough', 'bills', 'bathroom_counter', 'food_bacon', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'stereo', 'window', 'pillow', 'nightstand', 'bench', 'bed', 'clothes_gloves', 'keys', 'television', 'food_rice', 'lighter', 'cd', 'chessboard', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room. Turn on radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'character', 'shoe-shine_kit', 'faucet', 'couch', 'form', 'wall', 'bench', 'toy', 'cup', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom, find sink, turn on faucet, put soap in hands, wash and rinse hands,,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <soap>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'brush', 'ceilinglamp', 'stereo', 'electric_shaver', 'mousepad', 'closetdrawer', 'food_orange', 'wall', 'powersocket', 'photoframe', 'television', 'wine', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "walk into Living room, look for computer, walk to computer, turn on computer, grab mouse, click mouse", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[WALK] <computer>", "[TURNTO] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'beer', 'tea', 'freezer', 'shampoo', 'character', 'desk', 'piano_bench', 'ceilinglamp', 'iron', 'picture', 'mousepad', 'couch', 'bowl', 'nightstand', 'wall_clock', 'toy', 'bed', 'coffe_maker', 'television', 'coin', 'after_shave', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'shoes', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": take remote control, push \"change channel\" button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'crayon', 'drawing', 'hanger', 'floor', 'character', 'bedroom', 'mat', 'window', 'towel', 'pillow', 'wall', 'nightstand', 'cutting_board', 'cup', 'bed', 'toilet_paper', 'food_donut', 'curtain', 'doorjamb', 'mouse', 'shoes', 'chessboard', 'controller', 'laundry_detergent', 'door', 'table', 'vase']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "go t the bedroom and go to sleep", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'jelly', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'cleaning_solution', 'toaster', 'mat', 'faucet', 'window', 'couch', 'bowl', 'wall', 'oven', 'cup', 'food_food', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'instrument_violin', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Washing hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[SCRUB] <hands_both>', '[SWITCHON] <faucet>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'beer', 'toilet', 'cat', 'orchid', 'wallshelf', 'floor', 'desk', 'mat', 'deck_of_cards', 'window', 'couch', 'candle', 'pillow', 'towel', 'wall', 'wall_clock', 'powersocket', 'check', 'coffe_maker', 'television', 'food_food', 'curtain', 'doorjamb', 'vacuum_cleaner', 'controller', 'instrument_guitar', 'table', 'light', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to living room and lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'band-aids', 'bathroom', 'cat', 'orchid', 'wallshelf', 'hanger', 'floor', 'phone', 'character', 'bedroom', 'purse', 'computer', 'table_cloth', 'tvstand', 'powersocket', 'toy', 'clothes_gloves', 'coffe_maker', 'razor', 'food_onion', 'centerpiece', 'headset', 'chessboard', 'cupboard', 'door', 'home_office', 'table', 'laser_pointer']\nTask:Open door\nresponse:\n{\n    \"thoughts\": I walk up to the door.  I use my hand to turn the door knob to the left.  I pull the door open with the same hand I used to turn the door knob,\n    \"actions\": ['[RUN] <home_office>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'food_ice_cream', 'toothbrush', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'purse', 'faucet', 'window', 'towel', 'wall', 'toy', 'bed', 'light_bulb', 'novel', 'coffe_maker', 'television', 'water', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'dresser', 'maindoor', 'sink', 'tooth_paste', 'clothes_shirt', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[FIND] <toothbrush>", "[FIND] <water>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'ceilinglamp', 'floor_lamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'oven', 'photoframe', 'novel', 'television', 'doorjamb', 'dresser', 'cupboard', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'food_butter', 'knife', 'wallshelf', 'bathroom_counter', 'phone', 'character', 'bedroom', 'keyboard', 'computer', 'faucet', 'electric_shaver', 'mousepad', 'wall', 'powersocket', 'oven', 'shelf', 'cup', 'bed', 'food_rice', 'tray', 'mouse', 'headset', 'comb', 'dresser', 'conditioner', 'home_office', 'table']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "go to the bathroom and brush your teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'bookshelf', 'clothes_jacket', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'shower', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and find the sofa. I sit on the sofa and get comfortable.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'freezer', 'detergent', 'tea_bag', 'food_bacon', 'mop_bucket', 'stereo', 'iron', 'couch', 'pasta', 'powersocket', 'oven', 'wall_clock', 'photoframe', 'coffe_maker', 'mouthwash', 'television', 'food_rice', 'tray', 'after_shave', 'towel_rack', 'kitchen_counter', 'cd', 'dishrack', 'comb', 'dresser', 'clothes_skirt', 'creditcard']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the radio.  I switch on the radio and listen", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Do homework\nresponse:\n{\n    \"thoughts\": I do my online homework on the computer.  I use Pearson My Math Lab for math class and excel for my engineering class.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'food_carrot', 'wallshelf', 'floor', 'ceiling', 'dvd_player', 'stovefan', 'shoe-shine_kit', 'shower', 'couch', 'wall', 'oven', 'powersocket', 'cutting_board', 'food_cheese', 'standingmirror', 'coffe_maker', 'food_food', 'water', 'drinking_glass', 'tray', 'after_shave', 'kitchen_counter', 'home_office', 'teeth']\nTask:Drink\nresponse:\n", "output": {"thoughts": "im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'beer', 'measuring_cup', 'toilet', 'drawing', 'knife', 'hanger', 'floor', 'desk', 'tea_bag', 'keyboard', 'piano_bench', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'foundation', 'clothes_pants', 'towel', 'wall', 'nightstand', 'novel', 'television', 'tray', 'curtain', 'bag', 'table', 'vase', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": BEFORE EATING,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <towel>', '[WIPE] <towel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'bathroom', 'knife', 'bathroom_counter', 'floor', 'shampoo', 'ceiling', 'stovefan', 'mat', 'faucet', 'mousepad', 'couch', 'food_peanut_butter', 'dustpan', 'towel', 'wall', 'wall_clock', 'oven', 'novel', 'ice', 'tray', 'rag', 'shoes', 'maindoor', 'sink', 'bathroom_cabinet', 'vase', 'soap', 'tooth_paste', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'juice', 'dog', 'freezer', 'walllamp', 'ground_coffee', 'knifeblock', 'floor', 'ceiling', 'tea_bag', 'envelope', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'couch', 'food_kiwi', 'wall', 'bed', 'newspaper', 'check', 'coffe_maker', 'food_food', 'curtain', 'kitchen_counter', 'cupboard', 'pajamas', 'food_snack', 'sink', 'coffee']\nTask:Have snack\nresponse:\n{\n    \"thoughts\": Pour soda into a glass. Get pretzels out of bag. Eat and drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTIN] <juice> <freezer>', '[CLOSE] <freezer>', '[WALK] <cupboard>', '[OPEN] <cupboard>', '[FIND] <food_snack>', '[GRAB] <food_snack>', '[CLOSE] <cupboard>', '[DRINK] <cup>', '[EAT] <food_snack>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'freezer', 'milk', 'floor', 'remote_control', 'thread', 'button', 'mat', 'stereo', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'tray', 'curtain', 'doorjamb', 'coffee_pot', 'dresser', 'hairbrush', 'sink', 'home_office', 'table', 'bathroom_cabinet', 'laser_pointer']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "go to living room. grab remote control. sit on couch. point remote at tv. push buttons.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'laptop', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watching TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_salt', 'tea', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'broom', 'wall', 'wall_clock', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'maindoor', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to my desk. I turn on the computer. I open a web page and look at it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'tablelamp', 'freezer', 'blow_dryer', 'trashcan', 'orchid', 'hanger', 'floor', 'computer', 'button', 'stovefan', 'ceilinglamp', 'window', 'couch', 'wall', 'nightstand', 'oven', 'powersocket', 'bed', 'novel', 'tray', 'curtain', 'after_shave', 'towel_rack', 'pencil', 'maindoor', 'sink', 'light', 'dishwasher']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Walk to bedroom. Walk to bed, lie in bed.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'toilet', 'needle', 'hanger', 'floor', 'ceiling', 'dvd_player', 'stamp', 'desk', 'toaster', 'mat', 'tvstand', 'stereo', 'stovefan', 'electric_shaver', 'ceilinglamp', 'clothes_hat', 'wall_clock', 'powersocket', 'toy', 'photoframe', 'oven_mitts', 'food_donut', 'tray', 'pencil', 'comb', 'door', 'home_office', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Go to the living room.  Find the telephone.  Lift up the receiver.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'orchid', 'hanger', 'ceiling', 'keyboard', 'piano_bench', 'tvstand', 'shower', 'electric_shaver', 'clothes_socks', 'couch', 'wall', 'powersocket', 'bench', 'cutting_board', 'toy', 'photoframe', 'bed', 'razor', 'wooden_spoon', 'after_shave', 'coffee_pot', 'dresser', 'clothes_dress', 'bathtub', 'bathroom_cabinet', 'light', 'clothes_shirt', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to kitchen. grab clothes_dress. look at washing machine. open washing machine. put clothes_dress on washing machine. close washing machine. switch on washing machine. switch off washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[TURNTO] <washing_machine>', '[LOOKAT] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>', '[SWITCHOFF] <washing_machine>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'shelf', 'novel', 'television', 'food_onion', 'doorjamb', 'kettle', 'mouse', 'kitchen_counter', 'dresser', 'clothes_dress', 'door', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "turn on the computer and browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'fax_machine', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Write report\nresponse:\n{\n    \"thoughts\": Sit at a desk, turn on a computer.  Open an existing document and revise it.  Print a hard copy for mailing.  Email a copy to the client.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[PULL] <mouse>', '[TOUCH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[TYPE] <keyboard>', '[TOUCH] <mouse>', '[FIND] <fax_machine>', '[SWITCHON] <fax_machine>', '[SWITCHOFF] <fax_machine>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'beer', 'chair', 'band-aids', 'freezer', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'towel', 'pillow', 'wall', 'powersocket', 'clothes_gloves', 'television', 'doorjamb', 'mouse', 'dresser', 'laundry_detergent', 'maindoor', 'bag', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to the Home Office.  Find the computer.  Turn on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'bathroom', 'detergent', 'trashcan', 'food_carrot', 'knifeblock', 'floor', 'bedroom', 'thread', 'computer', 'brush', 'plate', 'picture', 'window', 'mousepad', 'clothes_hat', 'pillow', 'wall', 'box', 'bed', 'television', 'curtain', 'food_onion', 'kitchen_counter', 'maindoor', 'table', 'light', 'clothes_shirt', 'dirt', 'teeth']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Climb into bed, then read the novel,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <bed>', '[LIE] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_salt', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'oil', 'wall', 'nightstand', 'powersocket', 'toy', 'bed', 'coffe_maker', 'television', 'doorjamb', 'centerpiece', 'mouse', 'dresser', 'hair', 'light', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to office, sit at desk, turn on computer, enter password, open application and begin work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[PUSH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'phone', 'character', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'shower', 'faucet', 'window', 'oil', 'couch', 'clothes_pants', 'pillow', 'wall', 'tape', 'cup', 'novel', 'tray', 'curtain', 'mouse', 'chessboard', 'controller', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on a couch, grab a book, open it, and scan the pages,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'cpuscreen', 'hanger', 'floor', 'bookshelf', 'bookmark', 'stovefan', 'mat', 'faucet', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'cutting_board', 'cup', 'newspaper', 'novel', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'hair', 'clothes_dress', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to the living room and relax on the sofa with a book", "actions": ["[RUN] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[DROP] <bookmark>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'ground_coffee', 'walllamp', 'dough', 'food_carrot', 'floor', 'ceiling', 'stovefan', 'toaster', 'stereo', 'clothes_hat', 'wall', 'oven', 'bench', 'cutting_board', 'food_cheese', 'coffe_maker', 'food_food', 'check', 'toothbrush_holder', 'food_donut', 'tray', 'kitchen_counter', 'dishrack', 'hairbrush', 'bag', 'bathroom_cabinet', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Open the fridge, put the groceries in the fridge and close it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'powersocket', 'check', 'television', 'wine', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table', 'clothes_skirt']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "enter office, find computer, turn it on,", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TOUCH] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'toaster', 'ceilinglamp', 'window', 'floor_lamp', 'couch', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'comb', 'door', 'bathtub', 'bathroom_cabinet', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Turning on lights,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sponge', 'mop', 'chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'tea_bag', 'bookshelf', 'computer', 'toaster', 'faucet', 'mousepad', 'closetdrawer', 'clothes_hat', 'wall', 'powersocket', 'bed', 'standingmirror', 'television', 'doorjamb', 'cards', 'mouse', 'dresser', 'hair', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.", "actions": ["[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'computer', 'food_bacon', 'mat', 'tvstand', 'ceilinglamp', 'plate', 'window', 'shower', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'wine', 'curtain', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'beer', 'tea', 'freezer', 'shampoo', 'character', 'desk', 'piano_bench', 'ceilinglamp', 'iron', 'picture', 'mousepad', 'couch', 'bowl', 'nightstand', 'wall_clock', 'toy', 'bed', 'coffe_maker', 'television', 'coin', 'after_shave', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'shoes', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "take remote control, push \"change channel\" button", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'walllamp', 'ground_coffee', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'check', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'bathroom_cabinet', 'table', 'scrabble', 'coffee', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Relax on the sofa in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'blender', 'toilet', 'detergent', 'knife', 'hanger', 'floor', 'character', 'food_cake', 'brush', 'mat', 'plate', 'faucet', 'dish_soap', 'oil', 'closetdrawer', 'food_jam', 'dining_room', 'food_kiwi', 'oven', 'powersocket', 'cup', 'after_shave', 'kitchen_counter', 'dishrack', 'pajamas', 'spoon', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find dishwasher. Open dishwasher. Find dishrack. Find plate 1. Grab plate 1. Put plate 1 in dishrack. Repeat for dish 2. Find dish soap. Grab dish soap. Put dish soap to dishwasher. Close dishwasher. Switch on dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'band-aids', 'walllamp', 'drawing', 'bills', 'orchid', 'floor', 'ceiling', 'character', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'plate', 'oil', 'mousepad', 'wall', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'doorjamb', 'door', 'napkin', 'table', 'teeth']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch TV in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'measuring_cup', 'microwave', 'chair', 'freezer', 'detergent', 'orchid', 'knife', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'floor', 'desk', 'mop_bucket', 'ceilinglamp', 'floor_lamp', 'food_jam', 'pillow', 'dining_room', 'photoframe', 'keys', 'toothbrush_holder', 'oven_mitts', 'wine', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'dirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into my bedroom, go to the lamp and use the switch to turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[LOOKAT] <floor_lamp>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'bills', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch  TV\nresponse:\n{\n    \"thoughts\": Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'dough', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookmark', 'stovefan', 'brush', 'deck_of_cards', 'faucet', 'window', 'closetdrawer', 'dining_room', 'wall', 'oven', 'cutting_board', 'cup', 'bed', 'coffe_maker', 'food_food', 'tray', 'shaving_cream', 'kitchen_counter', 'chessboard', 'clothes_dress', 'conditioner']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[TURNTO] <cup>", "[LOOKAT] <cup>", "[WALK] <cup>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[WALK] <freezer>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_pizza', 'mop', 'bathroom', 'toothbrush', 'drawing', 'knifeblock', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'mop_bucket', 'brush', 'toaster', 'faucet', 'stereo', 'couch', 'towel', 'wall', 'powersocket', 'cup', 'photoframe', 'bed', 'mouse', 'dresser', 'sink', 'bathroom_cabinet', 'soap', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Upon entering the restroom, walk to the sink. Take a toothbrush and apply enough toothpaste so that 1/4 of the brush is covered (this is the appropriate amount). Proceed to \"brush\" your teeth using an up and down motion with the brush in your mouth. Brush for 2 minutes covering all areas of the teeth. Rinse with water and clean toothbrush with water too.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[SCRUB] <tooth_paste>', '[FIND] <toothbrush>', '[WASH] <toothbrush>', '[FIND] <sink>', '[RINSE] <sink>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'stovefan', 'mat', 'faucet', 'oil', 'bowl', 'wall', 'wall_clock', 'oven', 'cutting_board', 'clothes_gloves', 'food_food', 'television', 'tray', 'food_onion', 'lighter', 'kitchen_counter', 'dishrack', 'sink', 'bathroom_cabinet', 'teeth']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I switch on the stove and cook my groceries there.", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[SWITCHON] <oven>", "[FIND] <bowl>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <oven>", "[FIND] <food_food>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'toilet', 'nail_polish', 'detergent', 'cat', 'trashcan', 'milk', 'knifeblock', 'phone', 'ceiling', 'desk', 'keyboard', 'mousepad', 'couch', 'pillow', 'bench', 'oven', 'cup', 'bed', 'light_bulb', 'food_food', 'lighter', 'kitchen_counter', 'hairbrush', 'dresser', 'laundry_detergent', 'table', 'clothes_skirt', 'tooth_paste', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Petting a cat may sound simple.Letting the Cat Come to You.Extend a hand or finger and allow the cat a chance to touch her nose to you. When a cat bumps her head into your hand, it's a signal she wants attention.Focusing on Areas with Scent Glands. Feed with food.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <milk>', '[FIND] <milk>', '[GRAB] <milk>', '[POUR] <milk> <cup>', '[TURNTO] <cat>', '[POINTAT] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'toilet', 'toothbrush', 'drawing', 'floor', 'desk', 'tea_bag', 'toaster', 'mat', 'deck_of_cards', 'window', 'oil', 'mousepad', 'couch', 'pillow', 'broom', 'wall', 'bench', 'cutting_board', 'standingmirror', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'lighter', 'maindoor', 'napkin', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Find the sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'food_ice_cream', 'hanger', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'faucet', 'ceilinglamp', 'wall', 'powersocket', 'photoframe', 'clothes_gloves', 'bed', 'food_food', 'check', 'ice', 'doorjamb', 'pencil', 'kitchen_counter', 'laundry_detergent', 'door', 'napkin', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on the light,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['fork', 'juice', 'chair', 'walllamp', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'powersocket', 'oven', 'television', 'coin', 'food_donut', 'doorjamb', 'mouse', 'dresser', 'maindoor', 'bathtub', 'light', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Turn on PC and log onto email account. Type email and send it. Switch off computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'tea', 'knifeblock', 'cpuscreen', 'folder', 'bedroom', 'table_cloth', 'toaster', 'plate', 'homework', 'oil', 'closetdrawer', 'powersocket', 'tape', 'bed', 'novel', 'coffe_maker', 'television', 'razor', 'wine', 'doorjamb', 'shoes', 'dishrack', 'headset', 'cupboard', 'hairdryer', 'light', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk into living room, walk to stereo, turn on stereo, listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'faucet', 'shower', 'oil', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'oven', 'novel', 'colander', 'tray', 'lighter', 'dresser', 'sink', 'bathroom_cabinet', 'light', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to bathroom, go to sink, turn the water on via faucet, grab soap, use the soap to scrub my hands, put the soap back and wash my hands. Then I stop the water with faucet.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'ground_coffee', 'detergent', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'stovefan', 'toaster', 'faucet', 'shower', 'closetdrawer', 'candle', 'dining_room', 'wall', 'oven', 'photoframe', 'bed', 'coffe_maker', 'coffee_filter', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'table']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Make coffee in the kitchen.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[PLUGIN] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'beer', 'freezer', 'walllamp', 'bathroom', 'bathroom_counter', 'character', 'bedroom', 'desk', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'clothes_pants', 'towel', 'bench', 'basket_for_clothes', 'bed', 'coin', 'food_donut', 'curtain', 'towel_rack', 'mouse', 'cd', 'comb', 'laundry_detergent', 'bathtub', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'jelly', 'bathroom', 'detergent', 'food_butter', 'drawing', 'needle', 'cpuscreen', 'bathroom_counter', 'character', 'stovefan', 'plate', 'window', 'bowl', 'wall', 'bench', 'cup', 'bed', 'coffe_maker', 'check', 'tray', 'towel_rack', 'kitchen_counter', 'vacuum_cleaner', 'pajamas', 'sink', 'light', 'dishwasher', 'teeth']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room. Turn on radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'microwave', 'freezer', 'walllamp', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'alcohol', 'stovefan', 'toaster', 'tvstand', 'couch', 'wall', 'oven', 'cutting_board', 'toy', 'food_cheese', 'novel', 'coffe_maker', 'food_food', 'television', 'tray', 'shaving_cream', 'cards', 'kitchen_counter', 'shoes', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "put away groceries in fridge. open the door, put on shelf, close the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'detergent', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'purse', 'toaster', 'mat', 'faucet', 'iron', 'mousepad', 'wall', 'powersocket', 'cup', 'coffe_maker', 'water', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'clothes_scarf', 'colander']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'walllamp', 'dough', 'orchid', 'cpuscreen', 'hanger', 'floor', 'keyboard', 'remote_control', 'food_cake', 'envelope', 'computer', 'button', 'food_bacon', 'ceilinglamp', 'stereo', 'homework', 'oven', 'tape', 'photoframe', 'coffee_table', 'keys', 'check', 'novel', 'light_bulb', 'wooden_spoon', 'curtain', 'cd', 'home_office']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_ice_cream', 'beer', 'crayon', 'chair', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'phone', 'desk', 'bedroom', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'cup', 'bed', 'keys', 'food_food', 'television', 'food_onion', 'doorjamb', 'mouse', 'home_office']\nTask:Lock door\nresponse:\n{\n    \"thoughts\": Turn the lock.,\n    \"actions\": ['[WALK] <bedroom>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'microwave', 'toilet', 'freezer', 'food_butter', 'knife', 'floor', 'table_cloth', 'toaster', 'plate', 'faucet', 'stereo', 'dish_soap', 'oil', 'wall', 'wall_clock', 'oven', 'powersocket', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'coffee_filter', 'wooden_spoon', 'tray', 'kitchen_counter', 'sink', 'dishwasher', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'floor', 'dvd_player', 'bedroom', 'bookshelf', 'food_cake', 'shower', 'picture', 'foundation', 'mousepad', 'closetdrawer', 'tooth_paste', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'shelf', 'food_food', 'doorjamb', 'coffee_pot', 'mouse', 'pencil', 'kitchen_counter', 'towel_rack', 'door', 'napkin', 'bathroom_cabinet', 'board_game']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I go to the bathroom. I locate the toilet. I walk to the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[WALK] <toilet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'food_salt', 'freezer', 'detergent', 'ceiling', 'stamp', 'desk', 'bedroom', 'mat', 'deck_of_cards', 'faucet', 'shower', 'window', 'homework', 'food_peanut_butter', 'microphone', 'bowl', 'oven', 'toy', 'bed', 'light_bulb', 'standingmirror', 'check', 'doorjamb', 'kitchen_counter', 'hairbrush', 'hairdryer', 'bathroom_cabinet', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Read a book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'freezer', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'television', 'ice', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'headset', 'cupboard', 'sink', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'mop', 'ground_coffee', 'cat', 'floor', 'ceiling', 'envelope', 'stovefan', 'mat', 'tvstand', 'stereo', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'cutting_board', 'check', 'coffe_maker', 'television', 'video_game_controller', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'chessboard', 'dresser', 'door', 'table', 'colander']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'ceiling', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'child', 'bench', 'coffe_maker', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'home_office', 'table', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on tv.  Sit with kids.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[FIND] <child>', '[GREET] <child>', '[FIND] <child>', '[GREET] <child>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'mop', 'measuring_cup', 'microwave', 'chair', 'freezer', 'ground_coffee', 'bathroom', 'floor', 'ceiling', 'dvd_player', 'bookshelf', 'toaster', 'stovefan', 'faucet', 'iron', 'wall', 'wall_clock', 'cutting_board', 'shelf', 'clothes_gloves', 'coffe_maker', 'coffee_filter', 'food_rice', 'curtain', 'kitchen_counter', 'controller', 'sink', 'clothes_shirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "You walk into the kitchen and turn to the coffee maker. You grab the coffee filter and put into the coffee machine, you also grab the ground coffee and put it into the coffee machine. You turn on the coffee machine.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'freezer', 'walllamp', 'trashcan', 'drawing', 'orchid', 'ceiling', 'phone', 'bedroom', 'stovefan', 'mat', 'tvstand', 'shower', 'faucet', 'iron', 'food_noodles', 'couch', 'towel', 'pillow', 'dining_room', 'wall_clock', 'bench', 'bed', 'television', 'after_shave', 'doorjamb', 'door', 'bathtub', 'spoon', 'bathroom_cabinet']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to home office 1. Walk over to phone 1 on table 1.  Grab phone 1.  lift phone 1.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'microwave', 'freezer', 'walllamp', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'phone', 'food_chicken', 'stovefan', 'mat', 'faucet', 'wall', 'wall_clock', 'oven', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'wooden_spoon', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'food_steak', 'sink', 'bathroom_cabinet']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into the kitchen. Open the fridge. Take out the chicken. Open the kitchen cabinet. Take out the frying pan. Put the chicken in the frying pan. Put the frying pan on the stove. Turn on the stove. Cook the chicken.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'water_glass', 'pot', 'spectacles', 'walllamp', 'food_carrot', 'kitchen_cabinet', 'shampoo', 'longboard', 'computer', 'bookmark', 'mat', 'stereo', 'shower', 'homework', 'oil', 'clothes_socks', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'water', 'food_onion', 'headset', 'cd', 'dresser', 'controller', 'sink', 'napkin', 'clothes_skirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Go to kitchen. Open the the cabinet. Locate the water glass. Pour water into glass. Drag the bottle close to mouth. Drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[CLOSE] <kitchen_cabinet>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'stovefan', 'mat', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'dresser', 'napkin', 'bathroom_cabinet', 'light', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I want to relax on the sofa to some relaxing music in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <stereo>", "[PLUGIN] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'bathroom', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'button', 'brush', 'mat', 'toaster', 'faucet', 'picture', 'iron', 'pillow', 'broom', 'wall', 'oven', 'keys', 'novel', 'coin', 'tray', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'soap', 'clothes_shirt', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[PUSH] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['fork', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'stamp', 'envelope', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'powersocket', 'standingmirror', 'television', 'curtain', 'doorjamb', 'mouse', 'bag', 'home_office', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to my living room and switch on the television, and lay on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'toaster', 'window', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'wall_clock', 'television', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'clothes_shirt']\nTask:Organize closet\nresponse:\n{\n    \"thoughts\": Organize the hanging clothes in my closet by color and type and get rid of clothes I don't wear.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[CLOSE] <dresser>', '[WALK] <basket_for_clothes>', '[PUTBACK] <clothes_shirt> <basket_for_clothes>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'scissors', 'microwave', 'milk', 'knife', 'bathroom_counter', 'ceiling', 'bedroom', 'toaster', 'mat', 'plate', 'tvstand', 'dish_soap', 'bowl', 'food_kiwi', 'toy', 'cup', 'standingmirror', 'bed', 'food_food', 'food_donut', 'cards', 'dresser', 'clothes_dress', 'napkin', 'bathroom_cabinet', 'dishwasher', 'vase', 'board_game']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Using dishwasher to wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'chair', 'trashcan', 'drawing', 'clothes_underwear', 'floor', 'phone', 'character', 'keyboard', 'mat', 'ceilinglamp', 'plate', 'tvstand', 'shower', 'faucet', 'mousepad', 'couch', 'wall', 'bench', 'powersocket', 'newspaper', 'bed', 'television', 'mouse', 'controller', 'table', 'light', 'instrument_violin', 'creditcard']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.,\n    \"actions\": ['[WALK] <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['crayon', 'needle', 'clothes_underwear', 'phone', 'desk', 'stamp', 'clothes_jacket', 'stovefan', 'mat', 'tvstand', 'window', 'towel', 'pillow', 'cutting_board', 'basket_for_clothes', 'cup', 'bed', 'photoframe', 'newspaper', 'food_food', 'mouthwash', 'television', 'wine', 'kettle', 'dresser', 'bathtub', 'bathroom_cabinet', 'table', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and find the basket of clothes. Put the soap and clothes into the washing machine. Turn on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <washing_machine>", "[TURNTO] <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'stamp', 'bookshelf', 'clothes_jacket', 'mop_bucket', 'toaster', 'mat', 'brush', 'shoe-shine_kit', 'laptop', 'couch', 'pillow', 'wall', 'oven', 'cup', 'food_cheese', 'photoframe', 'check', 'television', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'bathroom_cabinet']\nTask:Eat cheese\nresponse:\n{\n    \"thoughts\": Open fridge.  Get cheese.  Eat it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[EAT] <food_cheese>']\n}\n", "input": "\nScene_Objects:['chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'computer', 'tvstand', 'faucet', 'stereo', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'wall', 'powersocket', 'shelf', 'food_food', 'oven_mitts', 'curtain', 'doorjamb', 'coffee_pot', 'mouse', 'vacuum_cleaner', 'dresser', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "approach the light switch, raise hand to the switch, flip the switch", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'drawing', 'orchid', 'floor', 'folder', 'mop_bucket', 'ceilinglamp', 'stereo', 'picture', 'window', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'bench', 'powersocket', 'novel', 'coffe_maker', 'toilet_paper', 'toothbrush_holder', 'tray', 'curtain', 'lighter', 'headset', 'cupboard', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and find the sofa. I sit on the sofa and get comfortable.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'measuring_cup', 'drawing', 'floor', 'bookshelf', 'mat', 'faucet', 'stereo', 'window', 'closetdrawer', 'microphone', 'food_orange', 'pillow', 'wall', 'nightstand', 'bed', 'food_food', 'curtain', 'towel_rack', 'headset', 'dresser', 'hairbrush', 'laundry_detergent', 'home_office', 'table', 'light', 'vase', 'board_game', 'scrabble']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Walk to bedroom. Walk to bed, lie in bed.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'bathroom', 'knifeblock', 'cpuscreen', 'hanger', 'bathroom_counter', 'ceiling', 'folder', 'character', 'keyboard', 'bookshelf', 'envelope', 'computer', 'dining_room', 'wall', 'nightstand', 'cup', 'clothes_gloves', 'light_bulb', 'check', 'curtain', 'pencil', 'kitchen_counter', 'controller', 'sink', 'napkin', 'clothes_skirt', 'tooth_paste']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I turn on computer. I use keyboard to type. I open browser. I search.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['freezer', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'thread', 'computer', 'stovefan', 'mat', 'faucet', 'clothes_socks', 'clothes_hat', 'towel', 'wall', 'wall_clock', 'oven', 'shelf', 'bed', 'standingmirror', 'water', 'tray', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go into the bathroom and go over to the sink. Grab soap and turn on the water. Wet your hands and rub them together so that the soap lathers. Rub the soap all over your hands thoroughly. Rinse off the soap with water and turn off the sink. Dry your hands with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'brush', 'mat', 'tvstand', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'bag', 'light', 'dishwasher', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'toilet', 'ground_coffee', 'blow_dryer', 'orchid', 'wallshelf', 'bathroom_counter', 'character', 'desk', 'bookshelf', 'alcohol', 'computer', 'ceilinglamp', 'laptop', 'faucet', 'pillow', 'wall_clock', 'powersocket', 'oven', 'cutting_board', 'cup', 'television', 'oven_mitts', 'after_shave', 'doorjamb', 'diary', 'towel_rack', 'cards', 'maindoor']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find cat. Walk to cat. Pull cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[PULL] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'newspaper', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'door', 'napkin', 'spoon', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch and use the remote control to change channels,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'chair', 'bathroom', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'character', 'computer', 'mail', 'table_cloth', 'toaster', 'foundation', 'window', 'mousepad', 'bench', 'shelf', 'photoframe', 'bed', 'coffe_maker', 'cards', 'chessboard', 'controller', 'door', 'maindoor', 'bathtub', 'light', 'vase', 'laser_pointer']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into living room, look at the phone when it ringing, grab the phone or pickup the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'band-aids', 'bathroom', 'trashcan', 'phone', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'picture', 'window', 'couch', 'bowl', 'broom', 'wall', 'oven', 'bench', 'cup', 'food_food', 'oven_mitts', 'colander', 'curtain', 'pencil', 'kitchen_counter', 'clothes_dress', 'sink', 'dishwasher', 'soap', 'instrument_violin']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find Cat in living room.  Walk to cat.  Pet cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_ice_cream', 'microwave', 'freezer', 'ground_coffee', 'nail_polish', 'milk', 'hanger', 'floor', 'ceiling', 'toaster', 'ceilinglamp', 'faucet', 'clothes_hat', 'wall', 'wall_clock', 'oven', 'cutting_board', 'shelf', 'coffe_maker', 'check', 'television', 'coffee_filter', 'water', 'curtain', 'kitchen_counter', 'maindoor', 'sink', 'vase', 'soap']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find coffee maker. Find coffee filter. Put it in the coffee maker. Find ground coffee. Put it in the coffee maker. Find Water. Put it in the coffe maker. Switch on the coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <water>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'walllamp', 'bathroom', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'stovefan', 'ceilinglamp', 'plate', 'shower', 'tvstand', 'food_noodles', 'bowl', 'wall', 'tape', 'shelf', 'bed', 'check', 'doorjamb', 'kitchen_counter', 'shoes', 'sink', 'light', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": walk to bathroom , find shower , take a shower,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'freezer', 'walllamp', 'food_carrot', 'desk', 'food_cake', 'remote_control', 'button', 'ceilinglamp', 'faucet', 'stereo', 'window', 'mousepad', 'couch', 'wall_clock', 'bench', 'box', 'cup', 'bed', 'keys', 'newspaper', 'food_food', 'ice', 'after_shave', 'maindoor', 'bathroom_cabinet', 'light', 'vase', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[PUSH] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'band-aids', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'homework', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'keys', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'light', 'soap']\nTask:Change TV channels\nresponse:\n{\n    \"thoughts\": Take remote. Change to desired channel by you.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[TOUCH] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'freezer', 'bathroom', 'cat', 'cpuscreen', 'bathroom_counter', 'bedroom', 'food_cake', 'computer', 'alcohol', 'toaster', 'stovefan', 'ceilinglamp', 'mousepad', 'closetdrawer', 'clothes_hat', 'nightstand', 'wall_clock', 'newspaper', 'coffee_filter', 'oven_mitts', 'kettle', 'kitchen_counter', 'dresser', 'hair', 'sink', 'napkin', 'bag', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find the cat, go near the cat and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'bathroom_counter', 'floor', 'mop_bucket', 'table_cloth', 'toaster', 'stovefan', 'plate', 'microphone', 'clothes_pants', 'dining_room', 'wall', 'pasta', 'cup', 'coffe_maker', 'food_food', 'mouse', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'home_office', 'light', 'dishwasher', 'colander', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <plate>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'detergent', 'cat', 'milk', 'phone', 'ceiling', 'desk', 'mop_bucket', 'brush', 'ceilinglamp', 'tvstand', 'faucet', 'toaster', 'mat', 'closetdrawer', 'microphone', 'nightstand', 'wall_clock', 'toy', 'cup', 'food_food', 'after_shave', 'lighter', 'towel_rack', 'mouse', 'pajamas', 'controller', 'laundry_detergent', 'soap']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go to the living room where the cat sleeps. Find cat. Proceed to petting cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'beer', 'chair', 'freezer', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mousepad', 'food_jam', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'food_onion', 'doorjamb', 'coffee_pot', 'mouse', 'kitchen_counter', 'dresser', 'door', 'table', 'light', 'dishwasher']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk to the bedroom, then find the computer and sit in the chair ,next switch on the computer and finally type in the keyboard to start to work on it,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'food_butter', 'floor', 'ceiling', 'stovefan', 'faucet', 'dustpan', 'food_jam', 'wall', 'oven', 'cutting_board', 'newspaper', 'coffe_maker', 'water', 'drinking_glass', 'tray', 'doorjamb', 'kitchen_counter', 'vacuum_cleaner', 'pajamas', 'home_office', 'bathroom_cabinet', 'clothes_skirt', 'vase', 'scrabble']\nTask:Drink\nresponse:\n", "output": {"thoughts": "go to the kitchen", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <water>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'bathroom', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'knifeblock', 'floor', 'dvd_player', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'faucet', 'electric_shaver', 'floor_lamp', 'mousepad', 'wall', 'powersocket', 'novel', 'food_food', 'check', 'television', 'after_shave', 'cupboard', 'sink', 'home_office', 'table', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Twist the lamp switch until the light goes on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'microwave', 'chair', 'freezer', 'ground_coffee', 'floor', 'ceiling', 'mail', 'food_bacon', 'toaster', 'deck_of_cards', 'faucet', 'window', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toy', 'light_bulb', 'coffe_maker', 'coffee_filter', 'water', 'curtain', 'after_shave', 'pencil', 'kitchen_counter', 'pajamas', 'sink', 'light']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'cat', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'ceilinglamp', 'homework', 'mousepad', 'wall', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'shoes', 'comb', 'dresser', 'hairdryer']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I open my laptop and then login to Windows. Then I open Google Chrome, and type gmail.com into the address bar. My email automatically logs in, then I click the Compose button, type in the address of the recipient, then the subject of the email, and then start typing my email,\n    \"actions\": ['[WALK] <laptop>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['glue', 'blender', 'microwave', 'toilet', 'bathroom', 'food_carrot', 'milk', 'wallshelf', 'cpuscreen', 'toaster', 'window', 'clothes_socks', 'pillow', 'wall', 'standingmirror', 'food_food', 'mouthwash', 'razor', 'ice', 'rag', 'lighter', 'cards', 'mouse', 'cd', 'hairbrush', 'pajamas', 'dresser', 'laundry_detergent', 'sink', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "walk to living room ,find a book ,open the book ,read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'jelly', 'chair', 'blow_dryer', 'bathroom', 'wallshelf', 'shampoo', 'phone', 'desk', 'clothes_jacket', 'stovefan', 'faucet', 'oil', 'closetdrawer', 'food_noodles', 'mousepad', 'microphone', 'towel', 'broom', 'bed', 'newspaper', 'after_shave', 'centerpiece', 'console', 'dresser', 'bathroom_cabinet', 'soap', 'tooth_paste', 'scrabble']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to living room. Find book. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'microwave', 'chair', 'food_butter', 'drawing', 'bathroom_counter', 'floor', 'stamp', 'desk', 'bookmark', 'stovefan', 'tvstand', 'shower', 'microphone', 'towel', 'dining_room', 'bowl', 'nightstand', 'oven', 'bench', 'shelf', 'food_cheese', 'photoframe', 'light_bulb', 'tray', 'kitchen_counter', 'dresser', 'hair', 'door', 'bathtub']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go into the living room and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'band-aids', 'drawing', 'floor', 'ceiling', 'desk', 'bookshelf', 'envelope', 'button', 'toaster', 'mat', 'clothes_hat', 'wall', 'powersocket', 'oven', 'toy', 'novel', 'toothbrush_holder', 'after_shave', 'doorjamb', 'pencil', 'door', 'light', 'table', 'dishwasher', 'colander']\nTask:Put toys away\nresponse:\n{\n    \"thoughts\": Pick up all the toys on the floor and put them in their correct storage bin or shelf.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['fork', 'scissors', 'microwave', 'bathroom', 'nail_polish', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'floor', 'bookshelf', 'mat', 'faucet', 'window', 'electric_shaver', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'photoframe', 'newspaper', 'keys', 'television', 'curtain', 'doorjamb', 'diary', 'sink', 'spoon', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'chair', 'walllamp', 'drawing', 'needle', 'wallshelf', 'bills', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'form', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'check', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'tooth_paste']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I enter the living room and sit on the couch. I grab the remote and turn on the TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'walllamp', 'drawing', 'orchid', 'bills', 'floor', 'desk', 'mat', 'tvstand', 'stereo', 'clothes_socks', 'closetdrawer', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'basket_for_clothes', 'photoframe', 'newspaper', 'clothes_gloves', 'standingmirror', 'mouthwash', 'television', 'razor', 'towel_rack', 'kitchen_counter', 'clothes_dress', 'sink', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "put clothes and soap in the washing machine, turn it on", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'kettle', 'mouse', 'dresser', 'hair', 'door', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": walk to living room. look at computer. switch on computer. sit in chair. watch computer. switch off computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[WATCH] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'button', 'food_bacon', 'stovefan', 'mat', 'shower', 'faucet', 'closetdrawer', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'keys', 'tray', 'doorjamb', 'lighter', 'sink', 'bathtub', 'bathroom_cabinet', 'soap']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "I want to take a shower so I go to the bathroom, turn on the shower, get the soap, lather and wash, then I rinse, turn off shower, get a towel and dry off.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[RINSE] <shower>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'freezer', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stovefan', 'mat', 'tvstand', 'faucet', 'toaster', 'teeth', 'towel', 'wall', 'wall_clock', 'oven', 'powersocket', 'food_food', 'tray', 'headset', 'comb', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'instrument_guitar', 'soap', 'hands_both', 'creditcard']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to bathroom, turn on faucet, wet hands, apply soap, rub hands, rinse hands, turn off faucet, dry hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'bathroom_counter', 'phone', 'tea_bag', 'keyboard', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'brush', 'foundation', 'window', 'closetdrawer', 'couch', 'broom', 'wall_clock', 'toy', 'food_cheese', 'novel', 'food_donut', 'tray', 'kitchen_counter', 'dresser', 'bathtub', 'light', 'shoe_rack']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the living room, Find Phone stand. Find Phone. Pickup the receiver to attend the call.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'homework', 'towel', 'dining_room', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bag', 'napkin', 'soap']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen counter, grab a grocery, walk to fridge, open fridge, put the grocery into fridge, and close fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'bathroom', 'bathroom_counter', 'floor', 'clothes_jacket', 'toaster', 'mat', 'tvstand', 'window', 'couch', 'candle', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'pasta', 'tape', 'newspaper', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'diary', 'pencil', 'hairbrush', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the sofa. Sit down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'food_bread', 'mop', 'freezer', 'trashcan', 'knifeblock', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'stovefan', 'mat', 'faucet', 'pillow', 'wall', 'powersocket', 'bed', 'food_food', 'check', 'mouthwash', 'after_shave', 'doorjamb', 'chessboard', 'dresser', 'cupboard', 'hair', 'door', 'instrument_guitar', 'light', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into living room. Find wall. Find light button. Switch on light button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'freezer', 'detergent', 'hanger', 'bathroom_counter', 'shampoo', 'ceiling', 'desk', 'bedroom', 'computer', 'stovefan', 'toaster', 'tvstand', 'shower', 'homework', 'closetdrawer', 'bench', 'oven', 'toy', 'cup', 'bed', 'photoframe', 'toilet_paper', 'coffe_maker', 'oven_mitts', 'pencil', 'dresser', 'bag', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'jelly', 'walllamp', 'bathroom', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'couch', 'food_jam', 'wall', 'novel', 'television', 'curtain', 'doorjamb', 'comb', 'clothes_dress', 'sink', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['mop', 'chair', 'bathroom', 'cpuscreen', 'hanger', 'computer', 'brush', 'toaster', 'mat', 'faucet', 'shower', 'dish_soap', 'food_noodles', 'mousepad', 'bowl', 'photoframe', 'standingmirror', 'novel', 'coffe_maker', 'check', 'television', 'bed', 'coffee_filter', 'curtain', 'clothes_dress', 'food_steak', 'door', 'spoon', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk to kitchen, find the dishwasher, apply dish soap into vessels or dishwasher, wash the dishes with a dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[WASH] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'tablelamp', 'blender', 'beer', 'wallshelf', 'floor', 'phone', 'stamp', 'desk', 'stovefan', 'mat', 'laptop', 'stereo', 'window', 'food_noodles', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'cup', 'photoframe', 'clothes_gloves', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'controller', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_ice_cream', 'chair', 'freezer', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'thread', 'computer', 'alcohol', 'mat', 'window', 'foundation', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'light', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and turn on the light. I walk to the couch then sit on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'blender', 'food_ice_cream', 'food_butter', 'phone', 'character', 'bookshelf', 'brush', 'ceilinglamp', 'plate', 'shower', 'iron', 'stereo', 'toaster', 'oil', 'food_orange', 'pillow', 'cup', 'food_food', 'check', 'oven_mitts', 'food_donut', 'food_onion', 'after_shave', 'mouse', 'hairdryer', 'bag', 'bathroom_cabinet', 'home_office']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <home_office>', '[DROP] <plate>', '[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <home_office>', '[DROP] <cup>', '[WALK] <dining_room>', '[WALK] <fork>', '[FIND] <fork>', '[GRAB] <fork>', '[WALK] <home_office>', '[DROP] <fork>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'bookshelf', 'brush', 'mat', 'stereo', 'picture', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'shelf', 'cup', 'television', 'wine', 'curtain', 'food_onion', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'home_office', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room. Walk to radio. Switch on the radio. Walk to sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'wallshelf', 'bathroom_counter', 'ceiling', 'desk', 'bookshelf', 'food_bacon', 'faucet', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'toilet_paper', 'check', 'razor', 'curtain', 'doorjamb', 'towel_rack', 'chessboard', 'shoes', 'cupboard', 'sink', 'bathroom_cabinet']\nTask:Get toilet paper\nresponse:\n{\n    \"thoughts\": There are extra toilet papers in the bottom cabinet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <toilet_paper>', '[GRAB] <toilet_paper>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['sheets', 'juice', 'freezer', 'walllamp', 'detergent', 'toothbrush', 'floor', 'ceiling', 'dvd_player', 'desk', 'table_cloth', 'closetdrawer', 'dining_room', 'wall', 'bench', 'cutting_board', 'light_bulb', 'check', 'mouthwash', 'video_game_controller', 'ice', 'curtain', 'doorjamb', 'coffee_pot', 'mouse', 'dishrack', 'hair', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "flip the switch and turn on the lights", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'stovefan', 'window', 'mat', 'plate', 'couch', 'candle', 'microphone', 'dining_room', 'wall', 'powersocket', 'box', 'photoframe', 'coffe_maker', 'television', 'curtain', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on the TV, sit on the couch, find a show, and watch it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'cd_player', 'knife', 'character', 'tea_bag', 'remote_control', 'button', 'table_cloth', 'ceilinglamp', 'shoe-shine_kit', 'shower', 'food_peanut_butter', 'candle', 'nightstand', 'bench', 'cutting_board', 'shelf', 'bed', 'photoframe', 'oven_mitts', 'food_rice', 'doorjamb', 'shaving_cream', 'mouse', 'dresser', 'controller', 'sink', 'table', 'clothes_skirt']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[PLUGIN] <cd_player>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'tablelamp', 'toilet', 'chair', 'walllamp', 'detergent', 'cat', 'knife', 'cpuscreen', 'floor', 'ceiling', 'mail', 'toaster', 'ceilinglamp', 'stovefan', 'stereo', 'homework', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'tape', 'newspaper', 'photoframe', 'food_food', 'television', 'water', 'tray', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": GOING TO DRINK WATER IN KITCHEN,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>', '[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'measuring_cup', 'freezer', 'bathroom', 'drawing', 'wallshelf', 'character', 'purse', 'envelope', 'ceilinglamp', 'nightstand', 'wall_clock', 'bench', 'cutting_board', 'newspaper', 'coffe_maker', 'tray', 'doorjamb', 'diary', 'cards', 'kitchen_counter', 'dresser', 'laundry_detergent', 'bathtub', 'light', 'table', 'tooth_paste', 'clothes_shirt', 'coffee']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'crayon', 'freezer', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'shampoo', 'desk', 'mail', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'stovefan', 'mat', 'electric_shaver', 'mousepad', 'clothes_pants', 'wall', 'nightstand', 'pasta', 'basket_for_clothes', 'food_food', 'diary', 'towel_rack', 'dresser', 'clothes_dress']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": in my home i walk to the kitchen with some groceries and locate the fridge,then open the fridge put the groceries into the fridge and closed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'table_cloth', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'wall_clock', 'oven', 'standingmirror', 'bed', 'food_food', 'water', 'tray', 'curtain', 'clothes_dress', 'sink', 'bathroom_cabinet', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I turn on faucet. I hold out my hands. I get my hands wet.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'button', 'food_bacon', 'stovefan', 'mat', 'shower', 'faucet', 'closetdrawer', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'keys', 'tray', 'doorjamb', 'lighter', 'sink', 'bathtub', 'bathroom_cabinet', 'soap']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": I want to take a shower so I go to the bathroom, turn on the shower, get the soap, lather and wash, then I rinse, turn off shower, get a towel and dry off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[RINSE] <shower>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'food_bread', 'microwave', 'toilet', 'freezer', 'food_carrot', 'dough', 'bathroom_counter', 'floor', 'phone', 'thread', 'stereo', 'shower', 'mousepad', 'food_jam', 'food_orange', 'form', 'bowl', 'food_kiwi', 'oven', 'shelf', 'coffe_maker', 'television', 'razor', 'towel_rack', 'coffee_pot', 'headset', 'chessboard', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to my bedroom, and then I switch on the radio", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to my home office. I find the computer along with mechanical keyboard on the desk. I sit on chair. I switch on the computer. I type on keyboard and work on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'jelly', 'bills', 'food_egg', 'hanger', 'floor', 'phone', 'ceiling', 'dvd_player', 'mat', 'window', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'wall', 'cutting_board', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'hair', 'table', 'soap', 'clothes_shirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to living room.  Find sofa.  Sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'detergent', 'dough', 'drawing', 'bills', 'clothes_underwear', 'floor', 'bedroom', 'keyboard', 'piano_bench', 'ceilinglamp', 'balanceball', 'picture', 'window', 'foundation', 'electric_shaver', 'couch', 'dustpan', 'food_kiwi', 'wall', 'keys', 'check', 'television', 'tray', 'rag', 'shoes', 'door', 'table', 'vase', 'coffee']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Relax on the sofa in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'faucet', 'shower', 'tooth_paste', 'towel', 'wall', 'wall_clock', 'oven', 'tape', 'photoframe', 'clothes_gloves', 'light_bulb', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'light', 'vase', 'soap', 'home_office', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom, open the faucet rinse the hands, wash the hands with soap and rinse again, then wipe with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[TURNTO] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_pizza', 'tablelamp', 'blender', 'food_salt', 'blow_dryer', 'trashcan', 'toothbrush', 'cpuscreen', 'hanger', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'purse', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'microphone', 'oven', 'food_food', 'check', 'tray', 'dresser', 'napkin', 'bathroom_cabinet', 'clothes_shirt']\nTask:Leave home\nresponse:\n{\n    \"thoughts\": Put on shoes. Put on coat. Open door and exit through it. Lock door behind you.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'centerpiece', 'mouse', 'shoes', 'dresser', 'door', 'maindoor', 'sink', 'light', 'table', 'soap']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "go to the office, turn computer on, open an browser or any application to open any website.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'ceilingfan', 'knifeblock', 'orchid', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'computer', 'couch', 'broom', 'wall', 'coffe_maker', 'video_game_controller', 'wine', 'doorjamb', 'towel_rack', 'cards', 'pencil', 'kitchen_counter', 'mouse', 'dresser', 'cupboard', 'door', 'light', 'board_game']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk over to light switch. Turn on lights.,\n    \"actions\": ['[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mail', 'button', 'clothes_jacket', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'keys', 'television', 'ice', 'doorjamb', 'mouse', 'pencil', 'headset', 'dresser', 'sink', 'light', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Switch on computer, Use keybord and mouse to open e-mail provider, type requested e-mail, click send, close e-mail provider, turn off computer", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[FIND] <keyboard>", "[FIND] <mail>", "[OPEN] <mail>", "[TYPE] <keyboard>", "[FIND] <button>", "[PUSH] <button>", "[CLOSE] <mail>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'ground_coffee', 'wallshelf', 'hanger', 'floor', 'stamp', 'desk', 'bookshelf', 'tea_bag', 'envelope', 'computer', 'mop_bucket', 'mat', 'shower', 'foundation', 'form', 'wall', 'powersocket', 'cutting_board', 'photoframe', 'novel', 'check', 'television', 'wooden_spoon', 'after_shave', 'doorjamb', 'door', 'light']\nTask:Read to child\nresponse:\n{\n    \"thoughts\": Read a book out loud to a child,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'tablelamp', 'mop', 'bathroom', 'drawing', 'floor', 'phone', 'character', 'desk', 'mail', 'laptop', 'tvstand', 'stereo', 'food_orange', 'dining_room', 'wall', 'oven', 'photoframe', 'television', 'tray', 'towel_rack', 'kettle', 'kitchen_counter', 'dresser', 'sink', 'bathtub', 'table', 'board_game', 'coffee']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'jelly', 'detergent', 'food_carrot', 'toothbrush', 'needle', 'hanger', 'floor', 'phone', 'shampoo', 'character', 'tvstand', 'iron', 'balanceball', 'stereo', 'mousepad', 'food_jam', 'pillow', 'nightstand', 'toothbrush_holder', 'diary', 'console', 'shaving_cream', 'kitchen_counter', 'sink', 'light', 'dishwasher', 'vase']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the playback device you want to listen to. turn on. press play. enjoyu music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'bills', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch  TV\nresponse:\n", "output": {"thoughts": "Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'floor', 'character', 'keyboard', 'toaster', 'mat', 'faucet', 'stereo', 'couch', 'towel', 'wall', 'wall_clock', 'nightstand', 'cup', 'coffe_maker', 'kitchen_counter', 'shoes', 'cd', 'hairbrush', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I have had my lunch and walked towards the sink and applied soap to my hands and washed cleanly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'walllamp', 'bathroom', 'trashcan', 'dough', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'faucet', 'window', 'food_noodles', 'mousepad', 'pillow', 'dining_room', 'wall', 'bed', 'food_food', 'razor', 'rag', 'console', 'mouse', 'towel_rack', 'kitchen_counter', 'door', 'sink', 'bathtub', 'bathroom_cabinet']\nTask:Clean sink\nresponse:\n", "output": {"thoughts": "Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.", "actions": ["[WALK] <bathroom>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[WIPE] <faucet>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'nail_polish', 'bathroom', 'detergent', 'knife', 'cpuscreen', 'dvd_player', 'bedroom', 'computer', 'table_cloth', 'mop_bucket', 'brush', 'toaster', 'shower', 'window', 'couch', 'dining_room', 'wall', 'pasta', 'cup', 'photoframe', 'after_shave', 'pencil', 'kitchen_counter', 'dresser', 'clothes_dress', 'sink', 'instrument_guitar', 'light', 'creditcard']\nTask:Sit\nresponse:\n{\n    \"thoughts\": We sit at the dining room table,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['pot', 'tea', 'microwave', 'bathroom', 'trashcan', 'orchid', 'knife', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'envelope', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'food_food', 'food_donut', 'curtain', 'doorjamb', 'kitchen_counter', 'hair', 'door', 'sink', 'table', 'light', 'dishwasher']\nTask:Put out flowers\nresponse:\n", "output": {"thoughts": "Place a flower centrepiece on the entrance hall table.", "actions": ["[WALK] <home_office>", "[WALK] <centerpiece>", "[FIND] <centerpiece>", "[GRAB] <centerpiece>", "[WALK] <table>", "[PUTBACK] <centerpiece> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'phone', 'character', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'shower', 'faucet', 'window', 'oil', 'couch', 'clothes_pants', 'pillow', 'wall', 'tape', 'cup', 'novel', 'tray', 'curtain', 'mouse', 'chessboard', 'controller', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on a couch, grab a book, open it, and scan the pages,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'toilet', 'chair', 'detergent', 'trashcan', 'orchid', 'floor', 'ceiling', 'stamp', 'man', 'bookshelf', 'purse', 'toaster', 'mat', 'oil', 'mousepad', 'nightstand', 'cup', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'curtain', 'mouse', 'headset', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Wait until doorbell rings, look to see who it is outside and open the door if you feel so inclined.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'stereo', 'window', 'shower', 'mousepad', 'wall', 'powersocket', 'bench', 'clothes_gloves', 'television', 'tray', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'bathtub', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk into home office. I walk to desk. I sit at chair. I turn on computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'microwave', 'chair', 'walllamp', 'bathroom', 'knifeblock', 'needle', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'plate', 'faucet', 'dish_soap', 'pillow', 'wall', 'oven', 'box', 'coffe_maker', 'coin', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'home_office', 'table']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Turn water on and run over dishes while scrubbing with a soapy sponge.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'juice', 'microwave', 'knifeblock', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'piano_bench', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'stovefan', 'tvstand', 'dish_soap', 'wall', 'pasta', 'coffee_filter', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'dishrack', 'cupboard', 'sink', 'scrabble']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <sink>', '[SWITCHOFF] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[GRAB] <plate>', '[SCRUB] <plate>', '[PUTOBJBACK] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[PUTOBJBACK] <plate>', '[SWITCHON] <faucet>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[GRAB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'food_salt', 'mop', 'microwave', 'band-aids', 'freezer', 'walllamp', 'bathroom', 'food_carrot', 'drawing', 'floor', 'ceiling', 'keyboard', 'bookshelf', 'computer', 'faucet', 'mousepad', 'couch', 'pillow', 'wall', 'photoframe', 'tray', 'curtain', 'doorjamb', 'shoes', 'door', 'bathroom_cabinet', 'light', 'shoe_rack']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'measuring_cup', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'doorjamb', 'mouse', 'pencil', 'cupboard', 'door', 'maindoor', 'home_office', 'light', 'dishwasher', 'creditcard']\nTask:Work\nresponse:\n{\n    \"thoughts\": turn computer on, open application to work with,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'toilet', 'band-aids', 'knifeblock', 'orchid', 'bedroom', 'keyboard', 'thread', 'clothes_jacket', 'faucet', 'food_noodles', 'food_orange', 'pillow', 'dining_room', 'oven', 'bench', 'powersocket', 'cup', 'newspaper', 'bed', 'check', 'coffe_maker', 'ice', 'tray', 'towel_rack', 'bathtub', 'home_office', 'dishwasher']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'mop', 'toilet', 'chair', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'computer', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'wall_clock', 'bench', 'television', 'wine', 'curtain', 'coffee_pot', 'door', 'sink', 'spoon', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I watch Netflix from about 8:30 to 10:30 every night. I usually watch TV shows.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <wall_clock>', '[FIND] <wall_clock>', '[TURNTO] <wall_clock>', '[LOOKAT] <wall_clock>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'microwave', 'walllamp', 'bathroom', 'detergent', 'knifeblock', 'knife', 'clothes_underwear', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'toaster', 'stovefan', 'plate', 'faucet', 'tvstand', 'wall', 'nightstand', 'oven', 'coffe_maker', 'drinking_glass', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'home_office', 'dishwasher']\nTask:Unload dishwasher\nresponse:\n", "output": {"thoughts": "Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[PUTBACK] <drinking_glass> <cupboard>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[PUTIN] <fork> <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTIN] <knife> <kitchen_cabinet>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTIN] <spoon> <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTIN] <plate> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'bedroom', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'pillow', 'dining_room', 'wall', 'bench', 'toy', 'box', 'photoframe', 'clothes_gloves', 'novel', 'video_game_controller', 'cards', 'vacuum_cleaner', 'dresser', 'controller', 'maindoor', 'napkin', 'dishwasher', 'instrument_violin']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'detergent', 'milk', 'knifeblock', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'computer', 'stovefan', 'toaster', 'tvstand', 'faucet', 'iron', 'couch', 'wall', 'oven', 'food_food', 'television', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'laser_pointer', 'scrabble']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Cooking a meal", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <pot> <sink>", "[GRAB] <pot>", "[SWITCHOFF] <faucet>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[CLOSE] <freezer>", "[WALK] <oven>", "[PUTBACK] <food_food> <pot>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door']\nTask:Watch youtube\nresponse:\n{\n    \"thoughts\": Go to youtube.Com watch videos,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'phone', 'bookshelf', 'table_cloth', 'bookmark', 'toaster', 'mat', 'faucet', 'oven', 'cup', 'photoframe', 'check', 'television', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'door', 'sink', 'light', 'tooth_paste']\nTask:Pet dog\nresponse:\n", "output": {"thoughts": "Pet my dog in hall", "actions": ["[WALK] <home_office>", "[WALK] <dog>", "[TURNTO] <dog>", "[LOOKAT] <dog>", "[TOUCH] <dog>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'freezer', 'cat', 'knifeblock', 'longboard', 'floor', 'ceiling', 'keyboard', 'clothes_jacket', 'stovefan', 'ceilinglamp', 'mat', 'stereo', 'tvstand', 'electric_shaver', 'mousepad', 'couch', 'wall', 'oven', 'bench', 'cup', 'food_food', 'doorjamb', 'conditioner', 'clothes_scarf', 'home_office', 'table', 'light', 'vase']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I get my groceries and go to the fridge. I open the fridge and put the groceries there.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'bathroom', 'knifeblock', 'bathroom_counter', 'shampoo', 'phone', 'mail', 'clothes_jacket', 'stovefan', 'ceilinglamp', 'window', 'clothes_socks', 'couch', 'clothes_pants', 'wall', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'television', 'mouse', 'shoes', 'chessboard', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'dishwasher', 'clothes_shirt']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "Open drawer. Pull out pants and put them on. Take a shirt out and put it on. Put on socks and shoes.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTON] <clothes_socks>", "[CLOSE] <dresser>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'television', 'tray', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch youtube\nresponse:\n{\n    \"thoughts\": Go to youtube.Com watch videos,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bedroom', 'bookshelf', 'keyboard', 'ceilinglamp', 'faucet', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'food_food', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathtub', 'light', 'table', 'instrument_violin']\nTask:Straighten pictures on wall\nresponse:\n", "output": {"thoughts": "Move the pictures/paintings until they are straight.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'food_salt', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'character', 'shower', 'stereo', 'microphone', 'wall', 'toy', 'cutting_board', 'cup', 'novel', 'food_food', 'check', 'food_onion', 'coffee_pot', 'console', 'shaving_cream', 'towel_rack', 'shoes', 'cd', 'hairbrush', 'maindoor', 'soap', 'coffee']\nTask:Read\nresponse:\n{\n    \"thoughts\": Read a book in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tea', 'toilet', 'walllamp', 'nail_polish', 'cpuscreen', 'floor', 'phone', 'ceiling', 'keyboard', 'toaster', 'mat', 'oil', 'pillow', 'bowl', 'wall', 'powersocket', 'oven', 'photoframe', 'television', 'rag', 'doorjamb', 'headset', 'door', 'sink', 'home_office', 'light', 'bathroom_cabinet', 'coffee']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'freezer', 'detergent', 'hanger', 'bathroom_counter', 'shampoo', 'ceiling', 'desk', 'bedroom', 'computer', 'stovefan', 'toaster', 'tvstand', 'shower', 'homework', 'closetdrawer', 'bench', 'oven', 'toy', 'cup', 'bed', 'photoframe', 'toilet_paper', 'coffe_maker', 'oven_mitts', 'pencil', 'dresser', 'bag', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['chair', 'trashcan', 'orchid', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'window', 'food_noodles', 'couch', 'pillow', 'wall', 'photoframe', 'toothbrush_holder', 'curtain', 'doorjamb', 'chessboard', 'door', 'bathtub', 'bag', 'home_office', 'light', 'table', 'dishwasher', 'instrument_violin']\nTask:Admire art\nresponse:\n", "output": {"thoughts": "Looking at painting, art", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'cat', 'dough', 'bathroom_counter', 'phone', 'bookshelf', 'ceilinglamp', 'tvstand', 'stereo', 'candle', 'bowl', 'dining_room', 'wall', 'newspaper', 'coffe_maker', 'food_food', 'coffee_filter', 'tray', 'curtain', 'diary', 'towel_rack', 'pencil', 'dresser', 'food_steak', 'sink', 'napkin', 'door', 'dishwasher']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[WALK] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'trashcan', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[PUTOBJBACK] <remote_control>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'cd_player', 'blow_dryer', 'bathroom', 'detergent', 'orchid', 'bills', 'bathroom_counter', 'phone', 'bedroom', 'remote_control', 'thread', 'button', 'mat', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'broom', 'wall_clock', 'powersocket', 'bench', 'box', 'coffe_maker', 'razor', 'towel_rack', 'vacuum_cleaner', 'bathtub', 'dishwasher', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[PLUGIN] <cd_player>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'shoes', 'door', 'table', 'light', 'creditcard']\nTask:Watch  TV\nresponse:\n", "output": {"thoughts": "Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'glue', 'filing_cabinet', 'microwave', 'jelly', 'freezer', 'ceilingfan', 'orchid', 'food_egg', 'floor', 'phone', 'ceiling', 'desk', 'thread', 'mat', 'laptop', 'deck_of_cards', 'faucet', 'foundation', 'pillow', 'nightstand', 'powersocket', 'basket_for_clothes', 'food_food', 'lighter', 'sink', 'home_office', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'bathroom', 'trashcan', 'orchid', 'drawing', 'wallshelf', 'floor', 'character', 'bedroom', 'tea_bag', 'keyboard', 'toaster', 'mat', 'stovefan', 'tvstand', 'couch', 'dining_room', 'wall', 'nightstand', 'child', 'bed', 'photoframe', 'novel', 'food_onion', 'after_shave', 'cupboard', 'bathtub', 'napkin', 'bathroom_cabinet', 'board_game']\nTask:Story reading time\nresponse:\n", "output": {"thoughts": "Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[SIT] <bed>", "[TURNTO] <child>", "[LOOKAT] <child>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'ceilingfan', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'food_kiwi', 'wall', 'oven', 'powersocket', 'cup', 'light_bulb', 'novel', 'television', 'tray', 'curtain', 'after_shave', 'kitchen_counter', 'dresser', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'centerpiece', 'shoes', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "TURN ON TV, SIT ON COUCH ACROSS THE ROOM, USE REMOTE TO LOOK FOR SHOW TO WATCH", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'bathroom', 'milk', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'envelope', 'computer', 'mat', 'ceilinglamp', 'stereo', 'mousepad', 'wall', 'toothbrush_holder', 'food_rice', 'ice', 'after_shave', 'mouse', 'headset', 'dresser', 'door', 'home_office', 'table', 'board_game']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office, walk to desk, sit in chair, turn on computer, look at computer, click mouse and type on keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'walllamp', 'dough', 'woman', 'knifeblock', 'bathroom_counter', 'floor', 'character', 'desk', 'tea_bag', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'wall', 'child', 'coffe_maker', 'video_game_controller', 'television', 'towel_rack', 'kettle', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'instrument_violin', 'clothes_shirt']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Go to the kitchen, greet your family members", "actions": ["[WALK] <dining_room>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'bedroom', 'computer', 'alcohol', 'mat', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'bench', 'cutting_board', 'standingmirror', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'chessboard', 'dresser', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit on the couch. Turn on TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'mop', 'walllamp', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'remote_control', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'powersocket', 'box', 'love_seat', 'photoframe', 'television', 'curtain', 'doorjamb', 'dresser', 'clothes_dress', 'bathtub', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I'll watch TV", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'ceilinglamp', 'shoe-shine_kit', 'faucet', 'clothes_pants', 'dining_room', 'wall', 'tape', 'basket_for_clothes', 'bed', 'coffe_maker', 'chessboard', 'clothes_dress', 'door', 'home_office', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'faucet', 'wall', 'cup', 'photoframe', 'keys', 'ice', 'tray', 'rag', 'towel_rack', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'dishwasher', 'table', 'laser_pointer']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'food_carrot', 'drawing', 'floor', 'thread', 'toaster', 'mat', 'picture', 'oil', 'towel', 'nightstand', 'powersocket', 'toy', 'shelf', 'box', 'cup', 'cutting_board', 'bed', 'photoframe', 'novel', 'toilet_paper', 'check', 'television', 'tray', 'food_onion', 'curtain', 'dresser', 'door', 'hairdryer', 'tooth_paste']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the radio; switch on the radio and listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "In the evenings after dinner, we sit on the couch, turn on the television, cue up Netflix and watch our favorite shows.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'balanceball', 'mousepad', 'candle', 'wall', 'nightstand', 'newspaper', 'doorjamb', 'mouse', 'light', 'clothes_skirt']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Turn on computer, pull up browser, browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'chair', 'freezer', 'drawing', 'knife', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'toaster', 'tvstand', 'shoe-shine_kit', 'faucet', 'window', 'mousepad', 'couch', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'curtain', 'coffee_pot', 'towel_rack', 'sink', 'light', 'laser_pointer']\nTask:Gaze out window\nresponse:\n", "output": {"thoughts": "Look at the yard from the window", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[FIND] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'walllamp', 'milk', 'floor', 'keyboard', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'faucet', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'toy', 'basket_for_clothes', 'coffe_maker', 'video_game_controller', 'tray', 'doorjamb', 'kettle', 'dresser', 'cupboard', 'clothes_dress', 'sink', 'table']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['tea', 'microwave', 'toilet', 'freezer', 'walllamp', 'ground_coffee', 'blow_dryer', 'knifeblock', 'cpuscreen', 'bedroom', 'computer', 'stovefan', 'faucet', 'mousepad', 'food_peanut_butter', 'pillow', 'wall', 'oven', 'love_seat', 'newspaper', 'television', 'after_shave', 'doorjamb', 'mouse', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'conditioner', 'bathtub']\nTask:Read\nresponse:\n", "output": {"thoughts": "Reading a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'measuring_cup', 'freezer', 'nail_polish', 'food_carrot', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'plate', 'mat', 'mousepad', 'food_peanut_butter', 'pillow', 'wall', 'powersocket', 'bench', 'food_vegetable', 'food_food', 'doorjamb', 'mouse', 'door', 'light', 'table', 'scrabble']\nTask:Fix snack\nresponse:\n{\n    \"thoughts\": Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTBACK] <food_carrot> <plate>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[PUTBACK] <food_vegetable> <plate>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'detergent', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'food_cake', 'computer', 'mat', 'faucet', 'mousepad', 'bowl', 'wall', 'nightstand', 'oven', 'bench', 'box', 'check', 'coffee_filter', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tea', 'microwave', 'bathroom', 'trashcan', 'orchid', 'knife', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'envelope', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'food_food', 'food_donut', 'curtain', 'doorjamb', 'kitchen_counter', 'hair', 'door', 'sink', 'table', 'light', 'dishwasher']\nTask:Put out flowers\nresponse:\n{\n    \"thoughts\": Place a flower centrepiece on the entrance hall table.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <centerpiece>', '[FIND] <centerpiece>', '[GRAB] <centerpiece>', '[WALK] <table>', '[PUTBACK] <centerpiece> <table>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'cards', 'cupboard', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on couch and watch the tv", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'dough', 'knifeblock', 'wallshelf', 'floor', 'phone', 'shampoo', 'desk', 'keyboard', 'bookmark', 'mat', 'shoe-shine_kit', 'floor_lamp', 'mousepad', 'pillow', 'powersocket', 'novel', 'curtain', 'centerpiece', 'kitchen_counter', 'door', 'bathtub', 'home_office', 'table', 'clothes_skirt', 'colander']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Turn a light on and read a book in the bedroom.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'bathroom', 'drawing', 'knifeblock', 'floor', 'character', 'keyboard', 'stovefan', 'mat', 'ceilinglamp', 'stereo', 'shower', 'window', 'toaster', 'wall', 'nightstand', 'box', 'food_cheese', 'bed', 'after_shave', 'lighter', 'diary', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'food_steak', 'light', 'dishwasher', 'clothes_shirt']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Turn on radio, put on favorite channel, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'tablelamp', 'detergent', 'cat', 'knife', 'drawing', 'hanger', 'bathroom_counter', 'desk', 'food_cake', 'purse', 'envelope', 'piano_bench', 'stovefan', 'toaster', 'stereo', 'balanceball', 'floor_lamp', 'powersocket', 'love_seat', 'newspaper', 'bed', 'photoframe', 'video_game_controller', 'diary', 'maindoor', 'light', 'soap', 'board_game']\nTask:Read\nresponse:\n{\n    \"thoughts\": Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'knifeblock', 'needle', 'wallshelf', 'bathroom_counter', 'phone', 'desk', 'food_cake', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'window', 'shower', 'stereo', 'food_peanut_butter', 'clothes_pants', 'pillow', 'bench', 'newspaper', 'bed', 'coffe_maker', 'television', 'lighter', 'dresser', 'cupboard', 'clothes_dress', 'table']\nTask:Get ready for school\nresponse:\n", "output": {"thoughts": "Take clothes out of dressers and get dressed for school.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[CLOSE] <dresser>", "[PUTON] <clothes_dress>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'jelly', 'walllamp', 'orchid', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'envelope', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'wall', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'bathroom_cabinet', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go into the office and sit down at the desk in front of the computer. Turn on the computer and look at it while clicking with the mouse and typing with the keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'newspaper', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'door', 'napkin', 'spoon', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch and use the remote control to change channels", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[PUTBACK] <remote_control> <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'toilet', 'chair', 'freezer', 'band-aids', 'bathroom', 'food_carrot', 'drawing', 'floor', 'mat', 'window', 'couch', 'clothes_hat', 'form', 'pillow', 'bench', 'food_food', 'check', 'coffe_maker', 'toilet_paper', 'ice', 'centerpiece', 'doorjamb', 'cupboard', 'home_office', 'table', 'light', 'bathroom_cabinet', 'scrabble']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'chair', 'food_carrot', 'cpuscreen', 'bathroom_counter', 'character', 'bedroom', 'keyboard', 'food_bacon', 'bookmark', 'stovefan', 'ceilinglamp', 'tvstand', 'picture', 'window', 'pillow', 'bench', 'love_seat', 'photoframe', 'novel', 'coin', 'curtain', 'towel_rack', 'mouse', 'pencil', 'sink', 'light', 'table', 'instrument_violin']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[READ] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[PUTBACK] <bookmark> <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'food_bacon', 'mat', 'ceilinglamp', 'shower', 'wall', 'wall_clock', 'cup', 'photoframe', 'bed', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'coffee']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[TURNTO] <cup>', '[LOOKAT] <cup>', '[WALK] <cup>', '[FIND] <freezer>', '[TURNTO] <freezer>', '[LOOKAT] <freezer>', '[WALK] <freezer>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'faucet', 'window', 'oil', 'couch', 'wall', 'bench', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'after_shave', 'towel_rack', 'kitchen_counter', 'bathtub', 'sink', 'light']\nTask:Open bathroom window\nresponse:\n", "output": {"thoughts": "Make sure bathroom window is unlocked. Lift up bathroom window so it is open.", "actions": ["[WALK] <bathroom>", "[WALK] <window>", "[FIND] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'oil', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'nightstand', 'wall', 'oven', 'bed', 'coffe_maker', 'television', 'curtain', 'dresser', 'door', 'instrument_guitar', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go to the living room, switch on the tv, take the remote and sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'knifeblock', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'television', 'curtain', 'doorjamb', 'cd', 'pajamas', 'door', 'table', 'light', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sitting on the couch watching my favorite television show.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'walllamp', 'milk', 'clothes_underwear', 'character', 'bedroom', 'keyboard', 'clothes_socks', 'candle', 'clothes_pants', 'wall', 'nightstand', 'powersocket', 'basket_for_clothes', 'photoframe', 'bed', 'food_food', 'curtain', 'diary', 'shoes', 'headset', 'laundry_detergent', 'sink', 'home_office', 'table', 'colander', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[DROP] <basket_for_clothes>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[GRAB] <laundry_detergent>', '[PUTOBJBACK] <laundry_detergent>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'tea', 'microwave', 'chair', 'freezer', 'walllamp', 'milk', 'knifeblock', 'knife', 'drawing', 'floor', 'ceiling', 'shampoo', 'desk', 'bookshelf', 'stovefan', 'toaster', 'faucet', 'food_jam', 'pillow', 'wall', 'oven', 'powersocket', 'cup', 'photoframe', 'tray', 'kitchen_counter', 'cupboard', 'controller', 'sink']\nTask:Get something to drink\nresponse:\n", "output": {"thoughts": "Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[CLOSE] <freezer>", "[POUR] <milk> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tea', 'chair', 'detergent', 'toothbrush', 'orchid', 'knifeblock', 'cpuscreen', 'kitchen_cabinet', 'floor', 'phone', 'bedroom', 'keyboard', 'ceilinglamp', 'faucet', 'window', 'dish_soap', 'food_jam', 'food_orange', 'pillow', 'wall', 'oven', 'bed', 'novel', 'coffe_maker', 'kitchen_counter', 'dishrack', 'cupboard', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTIN] <dish_soap> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'dog', 'bathroom', 'knife', 'phone', 'character', 'desk', 'clothes_jacket', 'mat', 'ceilinglamp', 'window', 'mousepad', 'pillow', 'dining_room', 'photoframe', 'keys', 'coffe_maker', 'food_food', 'check', 'ice', 'tray', 'towel_rack', 'mouse', 'shoes', 'cupboard', 'bathtub', 'home_office', 'light', 'bathroom_cabinet']\nTask:Get ready to leave\nresponse:\n", "output": {"thoughts": "Get your jacket and shoes on, locate the keys off the side table, walk out the front door and secure building with keys when you leave", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>", "[FIND] <keys>", "[GRAB] <keys>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'mop', 'dog', 'jelly', 'drawing', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'character', 'computer', 'stovefan', 'faucet', 'oil', 'dining_room', 'wall', 'nightstand', 'oven', 'cutting_board', 'photoframe', 'check', 'tray', 'sink', 'napkin', 'bathroom_cabinet', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": go to the bathroom and wash your hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[SQUEEZE] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'trashcan', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'keyboard', 'stovefan', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'newspaper', 'coffee_table', 'novel', 'bed', 'oven_mitts', 'doorjamb', 'pencil', 'chessboard', 'dresser', 'door', 'food_steak', 'table', 'light']\nTask:Read magazine\nresponse:\n", "output": {"thoughts": "Get magazine off coffee table. Sit in arm chair. Open magazine to first page. Read until finished.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <newspaper>", "[PUTBACK] <newspaper> <coffee_table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'chair', 'walllamp', 'bathroom', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'oven', 'newspaper', 'bed', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Use computer\nresponse:\n{\n    \"thoughts\": Sit at desk, use the computer mouse,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[TURNTO] <computer>', '[POINTAT] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'chair', 'bathroom', 'detergent', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'mat', 'tvstand', 'clothes_pants', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'novel', 'check', 'keys', 'tray', 'after_shave', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'tooth_paste']\nTask:Read\nresponse:\n", "output": {"thoughts": "I like to read books in my bedroom.", "actions": ["[WALK] <bedroom>", "[WALK] <desk>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'microwave', 'chair', 'wallshelf', 'hanger', 'floor', 'phone', 'food_bacon', 'piano_bench', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'powersocket', 'television', 'coin', 'food_donut', 'ice', 'rag', 'pencil', 'shoes', 'dresser', 'clothes_dress', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Hang up jacket\nresponse:\n{\n    \"thoughts\": I put clothes on hangers and hang them in the closet,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[FIND] <hanger>', '[OPEN] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['crayon', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'cat', 'trashcan', 'orchid', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'shower', 'window', 'mousepad', 'microphone', 'dining_room', 'wall', 'pasta', 'bench', 'oven', 'newspaper', 'check', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'home_office', 'bathroom_cabinet']\nTask:Read newspaper\nresponse:\n", "output": {"thoughts": "Pull out dining room chair. Sit in dining room chair. Open newspaper to section you'd like to read. Read section. Keep reading all sections you'd like.", "actions": ["[WALK] <dining_room>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'measuring_cup', 'freezer', 'ceilingfan', 'cpuscreen', 'longboard', 'ceiling', 'character', 'bedroom', 'toaster', 'stereo', 'balanceball', 'oil', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'nightstand', 'photoframe', 'coffe_maker', 'rag', 'doorjamb', 'towel_rack', 'shoes', 'controller', 'napkin', 'bathroom_cabinet', 'laser_pointer', 'dirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Go to the living room.  Find the telephone.  Lift up the receiver.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['walllamp', 'knifeblock', 'knife', 'cpuscreen', 'ceiling', 'bookshelf', 'purse', 'face', 'mail', 'mat', 'deck_of_cards', 'tvstand', 'homework', 'pillow', 'dining_room', 'nightstand', 'powersocket', 'box', 'photoframe', 'bed', 'novel', 'check', 'coffe_maker', 'mirror', 'tray', 'hair', 'clothes_dress', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Check appearance in mirror\nresponse:\n", "output": {"thoughts": "Walk to full-length mirror. Look in mirror. Make sure everything about your appearance is adequate. Adjust clothes as needed.", "actions": ["[WALK] <bedroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>", "[FIND] <face>", "[TURNTO] <face>", "[LOOKAT] <face>", "[FIND] <hair>", "[TURNTO] <hair>", "[LOOKAT] <hair>", "[FIND] <clothes_dress>", "[TURNTO] <clothes_dress>", "[LOOKAT] <clothes_dress>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'dog', 'bathroom', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'desk', 'shower', 'electric_shaver', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'newspaper', 'bed', 'food_food', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'bathroom_cabinet', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and find the sofa. I sit on the sofa and get comfortable.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'knifeblock', 'drawing', 'cpuscreen', 'orchid', 'floor', 'phone', 'bedroom', 'keyboard', 'mat', 'tvstand', 'floor_lamp', 'teeth', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'ice', 'tray', 'after_shave', 'doorjamb', 'centerpiece', 'mouse', 'dresser', 'light', 'coffee', 'creditcard']\nTask:Turn on lights\nresponse:\n", "output": {"thoughts": "Walk to your lamp, switch it on, and lay on your bed", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'freezer', 'blow_dryer', 'trashcan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'bookmark', 'ceilinglamp', 'mat', 'window', 'couch', 'microphone', 'clothes_hat', 'pillow', 'wall', 'bench', 'cup', 'keys', 'television', 'towel_rack', 'cd', 'cupboard', 'conditioner', 'door', 'home_office', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room and relax on the sofa with a book,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[DROP] <bookmark>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'chair', 'freezer', 'cat', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'character', 'stovefan', 'mat', 'tvstand', 'stereo', 'iron', 'food_kiwi', 'powersocket', 'newspaper', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'conditioner', 'bathtub', 'home_office', 'electrical_outlet']\nTask:Turn on radio\nresponse:\n", "output": {"thoughts": "Go to the radio.  Switch on the power button.  Tune the station.  Set the volume.  Listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <electrical_outlet>", "[TURNTO] <electrical_outlet>", "[PLUGIN] <stereo>", "[SWITCHON] <electrical_outlet>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'shampoo', 'ceiling', 'folder', 'face', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'picture', 'oil', 'food_orange', 'dining_room', 'bench', 'box', 'newspaper', 'check', 'video_game_controller', 'mirror', 'curtain', 'towel_rack', 'hair', 'clothes_dress', 'laundry_detergent', 'bathroom_cabinet', 'table', 'clothes_skirt', 'vase', 'scrabble']\nTask:Check appearance in mirror\nresponse:\n{\n    \"thoughts\": Walk to full-length mirror. Look in mirror. Make sure everything about your appearance is adequate. Adjust clothes as needed.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <mirror>', '[FIND] <mirror>', '[TURNTO] <mirror>', '[LOOKAT] <mirror>', '[FIND] <face>', '[TURNTO] <face>', '[LOOKAT] <face>', '[FIND] <hair>', '[TURNTO] <hair>', '[LOOKAT] <hair>', '[FIND] <clothes_dress>', '[TURNTO] <clothes_dress>', '[LOOKAT] <clothes_dress>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'tablelamp', 'mop', 'microwave', 'toilet', 'band-aids', 'bathroom', 'trashcan', 'wallshelf', 'character', 'computer', 'mat', 'faucet', 'couch', 'dining_room', 'wall', 'newspaper', 'oven_mitts', 'tray', 'curtain', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Shut front door\nresponse:\n", "output": {"thoughts": "Close the storm door. Lock the storm door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'ground_coffee', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'food_cake', 'mat', 'tvstand', 'shoe-shine_kit', 'homework', 'oil', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'bench', 'oven', 'powersocket', 'toilet_paper', 'kitchen_counter', 'door', 'tooth_paste']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk over and pick up the phone,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'knifeblock', 'drawing', 'wallshelf', 'floor', 'desk', 'keyboard', 'mat', 'tvstand', 'shower', 'faucet', 'stereo', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'bathtub', 'table', 'dishwasher']\nTask:Put in chair\nresponse:\n", "output": {"thoughts": "Take a comfortable chair. Place it in the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[GRAB] <chair>", "[WALK] <home_office>", "[WALK] <floor>", "[FIND] <floor>", "[PUTBACK] <chair> <floor>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'desk', 'computer', 'ceilinglamp', 'shower', 'window', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'bed', 'food_food', 'mouse', 'vacuum_cleaner', 'bathtub', 'sink', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk into the living room, walk over to phone, grab the phone, put it to your ear .,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['walllamp', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'keys', 'television', 'curtain', 'food_onion', 'doorjamb', 'mouse', 'dresser', 'pajamas', 'sink', 'home_office', 'light', 'instrument_guitar']\nTask:Turn off TV\nresponse:\n", "output": {"thoughts": "Walk over to tv and press power button", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'blender', 'freezer', 'walllamp', 'dough', 'orchid', 'drawing', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'window', 'picture', 'couch', 'pillow', 'bowl', 'wall', 'powersocket', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'dishrack', 'door', 'sink', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on the TV, sit on the couch, find a show, and watch it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'blender', 'microwave', 'chair', 'walllamp', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'mail', 'food_bacon', 'mat', 'stereo', 'faucet', 'homework', 'shower', 'dining_room', 'wall', 'oven', 'keys', 'television', 'food_onion', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'laundry_detergent', 'bag', 'spoon', 'table']\nTask:Put away keys\nresponse:\n", "output": {"thoughts": "Return keys into a bag", "actions": ["[WALK] <home_office>", "[WALK] <keys>", "[FIND] <keys>", "[GRAB] <keys>", "[FIND] <bag>", "[OPEN] <bag>", "[PUTBACK] <keys> <bag>", "[CLOSE] <bag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'character', 'desk', 'bookshelf', 'mat', 'deck_of_cards', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'bed', 'coffee_filter', 'wine', 'curtain', 'doorjamb', 'kitchen_counter', 'pajamas', 'door', 'food_steak', 'bag', 'light', 'electrical_outlet']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the electric outlet. I switch on the light.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <electrical_outlet>', '[FIND] <electrical_outlet>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['spectacles', 'glue', 'tea', 'chair', 'freezer', 'walllamp', 'knifeblock', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'keyboard', 'bookshelf', 'food_bacon', 'tvstand', 'faucet', 'window', 'mousepad', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'toilet_paper', 'diary', 'mouse', 'chessboard', 'spoon', 'home_office', 'table', 'light']\nTask:Sit in chair\nresponse:\n", "output": {"thoughts": "I sit in a chair to rest and eat.", "actions": ["[WALK] <dining_room>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'microwave', 'freezer', 'drawing', 'phone', 'bookshelf', 'toaster', 'stovefan', 'tvstand', 'mousepad', 'couch', 'clothes_hat', 'food_orange', 'wall', 'photoframe', 'clothes_gloves', 'bed', 'check', 'television', 'food_rice', 'tray', 'doorjamb', 'dresser', 'hairdryer', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'door']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to my bedroom, and then I switch on the radio,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['walllamp', 'nail_polish', 'trashcan', 'milk', 'drawing', 'wallshelf', 'bathroom_counter', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'computer', 'faucet', 'pillow', 'wall', 'photoframe', 'toilet_paper', 'check', 'coffe_maker', 'wine', 'after_shave', 'towel_rack', 'comb', 'controller', 'sink', 'bathroom_cabinet', 'light', 'table', 'instrument_violin', 'teeth']\nTask:Get toilet paper\nresponse:\n", "output": {"thoughts": "There are extra toilet papers in the bottom cabinet.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <toilet_paper>", "[GRAB] <toilet_paper>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'remote_control', 'computer', 'button', 'toaster', 'mat', 'shower', 'food_peanut_butter', 'food_kiwi', 'nightstand', 'oven', 'photoframe', 'television', 'oven_mitts', 'tray', 'kettle', 'cards', 'pencil', 'kitchen_counter', 'towel_rack', 'bathtub', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk into living room. Find television remote control. Pick up remote control. Press button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'scissors', 'microwave', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'tvstand', 'faucet', 'dining_room', 'wall', 'bed', 'novel', 'food_food', 'mouthwash', 'rag', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'home_office', 'bathroom_cabinet', 'soap', 'shoe_rack']\nTask:Added meat to freezer\nresponse:\n", "output": {"thoughts": "After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'freezer', 'walllamp', 'bathroom', 'bills', 'needle', 'drawing', 'character', 'table_cloth', 'toaster', 'picture', 'floor_lamp', 'window', 'mousepad', 'microphone', 'food_kiwi', 'wall', 'wall_clock', 'newspaper', 'bed', 'television', 'tray', 'cupboard', 'controller', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[TOUCH] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'toilet', 'chair', 'freezer', 'drawing', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'bookshelf', 'keyboard', 'mat', 'shoe-shine_kit', 'window', 'foundation', 'cup', 'photoframe', 'check', 'coffe_maker', 'centerpiece', 'kitchen_counter', 'shoes', 'bathtub', 'sink', 'bathroom_cabinet', 'light', 'napkin', 'instrument_violin']\nTask:Look at mirror\nresponse:\n", "output": {"thoughts": "I look at my reflection to check my cleanliness.", "actions": ["[WALK] <bathroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'mop', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'candle', 'pillow', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'check', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'door', 'bathroom_cabinet', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I go to the living room, switch off the light, sit in the sofa and do nothing.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHOFF] <light>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'detergent', 'trashcan', 'food_carrot', 'knifeblock', 'drawing', 'cpuscreen', 'knife', 'bathroom_counter', 'desk', 'keyboard', 'bookshelf', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'mousepad', 'pillow', 'dining_room', 'oven', 'photoframe', 'food_food', 'kitchen_counter', 'dresser', 'sink', 'home_office']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Walk to the lamp.  Find the switch.  Click the switch off.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[SWITCHOFF] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'freezer', 'food_butter', 'drawing', 'food_egg', 'floor', 'ceiling', 'character', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'couch', 'wall', 'nightstand', 'bench', 'oven', 'cup', 'novel', 'ice', 'curtain', 'doorjamb', 'kettle', 'pencil', 'headset', 'sink', 'bathroom_cabinet', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[TURNTO] <cup>', '[LOOKAT] <cup>', '[WALK] <cup>', '[FIND] <freezer>', '[TURNTO] <freezer>', '[LOOKAT] <freezer>', '[WALK] <freezer>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'toaster', 'mat', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'bench', 'photoframe', 'coffe_maker', 'food_rice', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'dishwasher', 'home_office']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'bathroom', 'detergent', 'food_butter', 'drawing', 'wallshelf', 'floor', 'desk', 'bookshelf', 'plate', 'faucet', 'picture', 'iron', 'window', 'dish_soap', 'food_jam', 'wall', 'nightstand', 'toy', 'cutting_board', 'food_food', 'television', 'curtain', 'pajamas', 'conditioner', 'light', 'dishwasher', 'tooth_paste', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'bathroom', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'knifeblock', 'floor', 'dvd_player', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'faucet', 'electric_shaver', 'floor_lamp', 'mousepad', 'wall', 'powersocket', 'novel', 'food_food', 'check', 'television', 'after_shave', 'cupboard', 'sink', 'home_office', 'table', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Twist the lamp switch until the light goes on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'keys', 'television', 'curtain', 'food_onion', 'doorjamb', 'mouse', 'dresser', 'pajamas', 'sink', 'home_office', 'light', 'instrument_guitar']\nTask:Turn off TV\nresponse:\n{\n    \"thoughts\": Walk over to tv and press power button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'sheets', 'chair', 'freezer', 'walllamp', 'milk', 'knife', 'drawing', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'mat', 'stovefan', 'window', 'food_jam', 'pillow', 'wall', 'bed', 'photoframe', 'food_food', 'check', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'home_office', 'table']\nTask:Look in refrigerator\nresponse:\n", "output": {"thoughts": "Approach the refrigerator. Use the handle to open the refrigerator door. Look at each shelf of the refrigerator. Look in each refrigerator drawer. Close the refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[WALK] <freezer>", "[OPEN] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <food_food>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'juice', 'tea', 'toilet', 'freezer', 'walllamp', 'dough', 'drawing', 'food_egg', 'floor', 'dvd_player', 'character', 'alcohol', 'stovefan', 'mat', 'shower', 'closetdrawer', 'clothes_hat', 'food_jam', 'wall', 'powersocket', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'laundry_detergent', 'door', 'light', 'dishwasher', 'table']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[TURNTO] <toilet>', '[LOOKAT] <toilet>', '[WALK] <toilet>', '[TURNTO] <toilet>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'dog', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'computer', 'stovefan', 'faucet', 'oil', 'pillow', 'wall', 'wall_clock', 'oven', 'pasta', 'keys', 'check', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub', 'light']\nTask:Keep an eye on stove as something is cooking\nresponse:\n", "output": {"thoughts": "Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <pot>", "[TURNTO] <pot>", "[LOOKAT] <pot>", "[FIND] <wall_clock>", "[TURNTO] <wall_clock>", "[LOOKAT] <wall_clock>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'scissors', 'microwave', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'tvstand', 'faucet', 'dining_room', 'wall', 'bed', 'novel', 'food_food', 'mouthwash', 'rag', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'home_office', 'bathroom_cabinet', 'soap', 'shoe_rack']\nTask:Added meat to freezer\nresponse:\n{\n    \"thoughts\": After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'freezer', 'knifeblock', 'drawing', 'wallshelf', 'ceiling', 'phone', 'desk', 'keyboard', 'face', 'toaster', 'laptop', 'tvstand', 'shower', 'stereo', 'homework', 'oil', 'towel', 'oven', 'basket_for_clothes', 'tray', 'towel_rack', 'hair', 'cupboard', 'door', 'bathtub', 'clothes_shirt']\nTask:Rain welcome\nresponse:\n", "output": {"thoughts": "The robot would greet me and then hand me a towel.  It would make an appropriate quip about the weather while I dried off.  It would receive the wet towel when I was done.", "actions": ["[WALK] <bedroom>", "[WALK] <towel>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <face>", "[WIPE] <face>", "[FIND] <hair>", "[WIPE] <hair>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'needle', 'floor', 'phone', 'mail', 'mat', 'plate', 'faucet', 'stereo', 'couch', 'dustpan', 'dining_room', 'wall', 'bench', 'oven', 'toy', 'cup', 'newspaper', 'standingmirror', 'food_food', 'razor', 'console', 'towel_rack', 'bathtub', 'table']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'chair', 'bathroom', 'drawing', 'wallshelf', 'clothes_underwear', 'hanger', 'character', 'bookshelf', 'clothes_jacket', 'piano_bench', 'ceilinglamp', 'shower', 'mousepad', 'teeth', 'pillow', 'check', 'toothbrush_holder', 'coin', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'coffee']\nTask:Close door\nresponse:\n", "output": {"thoughts": "Push the door shut until it latches.", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'jelly', 'drawing', 'floor', 'bedroom', 'table_cloth', 'stovefan', 'mat', 'plate', 'window', 'couch', 'wall', 'nightstand', 'wall_clock', 'cutting_board', 'bed', 'novel', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'pencil', 'cd', 'clothes_dress', 'bathroom_cabinet', 'table', 'light', 'vase', 'laser_pointer']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I would go lie down in my bed and open the book and start reading.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'measuring_cup', 'microwave', 'chair', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'cleaning_solution', 'keyboard', 'mop_bucket', 'stovefan', 'faucet', 'couch', 'clothes_pants', 'coffe_maker', 'mirror', 'curtain', 'rag', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'dishwasher']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'microwave', 'chair', 'knifeblock', 'drawing', 'keyboard', 'button', 'stovefan', 'toaster', 'shower', 'window', 'dustpan', 'pillow', 'dining_room', 'nightstand', 'tape', 'novel', 'television', 'tray', 'curtain', 'after_shave', 'doorjamb', 'kitchen_counter', 'cupboard', 'clothes_dress', 'sink', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I pick up my phone. I touch it to navigate to the web browser. I can swipe and tap the screen to visit different websites. Then I read or watch the content within those websites.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[TOUCH] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'freezer', 'walllamp', 'bathroom', 'cpuscreen', 'bathroom_counter', 'ceiling', 'keyboard', 'computer', 'food_bacon', 'table_cloth', 'ceilinglamp', 'faucet', 'picture', 'clothes_socks', 'towel', 'dining_room', 'wall', 'pasta', 'bed', 'centerpiece', 'kitchen_counter', 'door', 'bathtub', 'home_office', 'light', 'table', 'colander']\nTask:Clean screen\nresponse:\n", "output": {"thoughts": "Use a soft rag to wipe the screen until there are no marks on it.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <computer>", "[WIPE] <computer>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'thread', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink', 'light', 'table', 'creditcard']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on Computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'chair', 'bathroom', 'food_carrot', 'woman', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'man', 'keyboard', 'mat', 'faucet', 'window', 'foundation', 'nightstand', 'powersocket', 'cup', 'coffe_maker', 'food_food', 'television', 'tray', 'towel_rack', 'cd', 'dresser', 'sink', 'clothes_scarf', 'light']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Stand next to the door. When the door opens from the outside, greet the guests.", "actions": ["[RUN] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_pizza', 'tablelamp', 'food_ice_cream', 'crayon', 'shampoo', 'keyboard', 'mail', 'bookmark', 'stovefan', 'faucet', 'pillow', 'dining_room', 'broom', 'wall', 'love_seat', 'photoframe', 'novel', 'food_food', 'wine', 'tray', 'after_shave', 'shaving_cream', 'kettle', 'headset', 'pajamas', 'laundry_detergent', 'door', 'bathroom_cabinet', 'shoe_rack']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['microwave', 'dog', 'freezer', 'walllamp', 'cat', 'knifeblock', 'bills', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'faucet', 'clothes_hat', 'broom', 'powersocket', 'bench', 'coffe_maker', 'food_food', 'television', 'mirror', 'rag', 'lighter', 'towel_rack', 'mouse', 'dresser', 'sink', 'table', 'clothes_skirt', 'shoe_rack']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray glass cleaner on mirror. Wipe with a clean cloth.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[WALK] <mirror>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'chair', 'walllamp', 'cat', 'wallshelf', 'floor', 'character', 'desk', 'tea_bag', 'toaster', 'plate', 'couch', 'dining_room', 'box', 'newspaper', 'bed', 'toilet_paper', 'curtain', 'coffee_pot', 'mouse', 'kettle', 'cupboard', 'controller', 'clothes_dress', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find Cat in living room.  Walk to cat.  Pet cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'tablelamp', 'chair', 'walllamp', 'trashcan', 'knifeblock', 'bathroom_counter', 'keyboard', 'computer', 'brush', 'tvstand', 'shower', 'faucet', 'mousepad', 'wall', 'bench', 'cup', 'bed', 'photoframe', 'clothes_gloves', 'tray', 'wooden_spoon', 'towel_rack', 'dresser', 'cupboard', 'door', 'bathtub', 'napkin', 'bathroom_cabinet']\nTask:Wash monitor\nresponse:\n", "output": {"thoughts": "Clean the grime off of my monitor screen when not in use", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>", "[FIND] <brush>", "[TURNTO] <brush>", "[LOOKAT] <brush>", "[GRAB] <brush>", "[WIPE] <computer>", "[DROP] <brush>", "[FIND] <napkin>", "[GRAB] <napkin>", "[WASH] <napkin>", "[WIPE] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'chair', 'bathroom', 'cat', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'clothes_underwear', 'phone', 'ceiling', 'dvd_player', 'desk', 'stovefan', 'faucet', 'shower', 'closetdrawer', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'bed', 'tray', 'mouse', 'dresser', 'home_office']\nTask:Keep cats out of room\nresponse:\n{\n    \"thoughts\": Stand in the doorway and create a loud noise so the cats will stay out.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'mat', 'plate', 'faucet', 'ceilinglamp', 'dish_soap', 'mousepad', 'couch', 'dining_room', 'wall', 'oven', 'bench', 'basket_for_clothes', 'towel_rack', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'light']\nTask:Wash dirty dishes\nresponse:\n", "output": {"thoughts": "Wet the sponge, apply soap to sponge, scrub dishes, rinse dishes, place on dish rack.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <sponge> <sink>", "[GRAB] <sponge>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'keys', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I need to send an email to my friend so I go into my office and sit down at the desk. I turn on the computer and type the email.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[TURNTO] <desk>', '[LOOKAT] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'microwave', 'toilet', 'freezer', 'bathroom', 'trashcan', 'knife', 'cpuscreen', 'ceiling', 'bedroom', 'bookshelf', 'brush', 'stovefan', 'toaster', 'window', 'couch', 'wall_clock', 'powersocket', 'toy', 'bed', 'television', 'ice', 'curtain', 'mouse', 'dresser', 'controller', 'bathroom_cabinet', 'shoe_rack']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTBACK] <toy> <dresser>", "[CLOSE] <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'window', 'electric_shaver', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": watching TV from sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[DROP] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'chair', 'knifeblock', 'knife', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'character', 'desk', 'bedroom', 'bookshelf', 'food_bacon', 'mat', 'shoe-shine_kit', 'food_jam', 'wall', 'powersocket', 'oven', 'photoframe', 'novel', 'mouthwash', 'coffee_filter', 'tray', 'curtain', 'vacuum_cleaner', 'dresser', 'bathtub', 'soap']\nTask:Read to child\nresponse:\n", "output": {"thoughts": "Read a book out loud to a child", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'table_cloth', 'ceilinglamp', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'shelf', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office', 'light', 'table', 'laser_pointer']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Find the computer.  Turn on the computer.  Open the e-mail program.  Create a new message.  Write the message.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'walllamp', 'milk', 'floor', 'keyboard', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'faucet', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'toy', 'basket_for_clothes', 'coffe_maker', 'video_game_controller', 'tray', 'doorjamb', 'kettle', 'dresser', 'cupboard', 'clothes_dress', 'sink', 'table']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Bend down and pick up the toys off of the floor.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[WALK] <dresser>", "[GRAB] <toy>", "[PUTIN] <toy> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'bookmark', 'stovefan', 'ceilinglamp', 'mat', 'mousepad', 'oil', 'food_peanut_butter', 'broom', 'wall', 'novel', 'diary', 'mouse', 'cupboard', 'clothes_dress', 'door', 'table', 'light', 'dirt']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Go to office. Turn on computer. Pull out chair. Sit in chair. Use keyboard and mouse.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[GRAB] <mouse>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'beer', 'crayon', 'freezer', 'walllamp', 'bathroom', 'detergent', 'wallshelf', 'floor', 'ceiling', 'purse', 'computer', 'mat', 'tvstand', 'shower', 'window', 'couch', 'food_peanut_butter', 'dining_room', 'wall', 'photoframe', 'bed', 'television', 'curtain', 'diary', 'cupboard', 'controller', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Wipe down baseboards please\nresponse:\n", "output": {"thoughts": "Use a rag to get all the dust off the baseboards please.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <wall>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'ground_coffee', 'food_butter', 'dough', 'drawing', 'floor', 'desk', 'bookmark', 'mat', 'faucet', 'floor_lamp', 'window', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'novel', 'television', 'curtain', 'doorjamb', 'chessboard', 'controller', 'maindoor', 'home_office', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": First I walk into the living room, I find the lamp and switch it on. Then I find a book, grab it, sit in the sofa, open the book and read it. When I'm finished, I find a bookmark, grab it and put it in the book. Then I close the book, put it back, and switch off the lamp.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>', '[SWITCHOFF] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'microwave', 'walllamp', 'cat', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'toaster', 'faucet', 'window', 'dish_soap', 'dining_room', 'wall', 'oven', 'cup', 'drinking_glass', 'curtain', 'mouse', 'kitchen_counter', 'cd', 'dishrack', 'shoes', 'dresser', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet', 'bathtub']\nTask:Clean dishes\nresponse:\n", "output": {"thoughts": "Clean the dishes from the sink.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[SCRUB] <drinking_glass>", "[RINSE] <drinking_glass>", "[FIND] <dishrack>", "[PUTBACK] <drinking_glass> <dishrack>", "[FIND] <cup>", "[GRAB] <cup>", "[SCRUB] <cup>", "[RINSE] <cup>", "[PUTBACK] <cup> <dishrack>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'tablelamp', 'microwave', 'dog', 'walllamp', 'trashcan', 'knifeblock', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'faucet', 'ceilinglamp', 'oil', 'wall', 'oven', 'coffe_maker', 'drinking_glass', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'bag', 'dishwasher', 'colander']\nTask:Unload dishwasher\nresponse:\n{\n    \"thoughts\": Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[PUTBACK] <drinking_glass> <cupboard>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <fork> <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTIN] <knife> <kitchen_cabinet>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTIN] <spoon> <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_ice_cream', 'toilet', 'walllamp', 'dough', 'drawing', 'knifeblock', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bedroom', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'basket_for_clothes', 'bed', 'food_food', 'television', 'doorjamb', 'mouse', 'pencil', 'dishrack', 'door', 'home_office', 'light', 'bathroom_cabinet']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Grasp light switch, flip light switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'tablelamp', 'juice', 'cd_player', 'band-aids', 'walllamp', 'cpuscreen', 'bathroom_counter', 'ceiling', 'desk', 'stovefan', 'stereo', 'homework', 'food_orange', 'nightstand', 'bench', 'bed', 'newspaper', 'photoframe', 'coffe_maker', 'check', 'food_onion', 'curtain', 'doorjamb', 'dresser', 'door', 'sink', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk in the living room, then I go the the stereo and the cd player and turn them on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['sponge', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'character', 'desk', 'bookmark', 'mat', 'mousepad', 'couch', 'food_orange', 'pillow', 'wall', 'wall_clock', 'bed', 'photoframe', 'coffe_maker', 'check', 'doorjamb', 'towel_rack', 'door', 'bathtub', 'home_office', 'table', 'light', 'instrument_guitar', 'scrabble']\nTask:Read magazine\nresponse:\n", "output": {"thoughts": "Please sit on the couch and read a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <newspaper>", "[LOOKAT] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_oatmeal', 'microwave', 'freezer', 'bathroom', 'nail_polish', 'dough', 'food_carrot', 'orchid', 'drawing', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'brush', 'mat', 'couch', 'pillow', 'food_cereal', 'wall', 'food_vegetable', 'food_food', 'television', 'console', 'cards', 'kitchen_counter', 'cupboard', 'hair', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_vegetable> <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTIN] <food_carrot> <freezer>', '[FIND] <food_cereal>', '[GRAB] <food_cereal>', '[PUTIN] <food_cereal> <freezer>', '[FIND] <food_oatmeal>', '[GRAB] <food_oatmeal>', '[PUTIN] <food_oatmeal> <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[FIND] <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'filing_cabinet', 'walllamp', 'orchid', 'bathroom_counter', 'floor', 'phone', 'character', 'desk', 'computer', 'brush', 'toaster', 'mat', 'shower', 'faucet', 'electric_shaver', 'mousepad', 'dining_room', 'nightstand', 'oven', 'toy', 'cup', 'food_onion', 'mouse', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Put toys away\nresponse:\n", "output": {"thoughts": "Pick up all the toys on the floor and put them in their correct storage bin or shelf.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'freezer', 'bathroom', 'cpuscreen', 'desk', 'tea_bag', 'alcohol', 'tvstand', 'oven', 'tape', 'cutting_board', 'cup', 'clothes_gloves', 'novel', 'toilet_paper', 'television', 'coffee_filter', 'wooden_spoon', 'mouse', 'cd', 'controller', 'maindoor', 'sink', 'home_office', 'light', 'dishwasher', 'soap', 'dirt', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room. Turn on radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'juice', 'food_ice_cream', 'food_oatmeal', 'microwave', 'chair', 'freezer', 'walllamp', 'ground_coffee', 'milk', 'knifeblock', 'drawing', 'knife', 'floor', 'ceiling', 'stovefan', 'toaster', 'mat', 'faucet', 'ceilinglamp', 'bowl', 'wall', 'oven', 'bench', 'cup', 'coffe_maker', 'food_donut', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Put oatmeal into a bowl.  Add milk.  Heat in a microwave.  Put pod of coffee into a coffee maker.  Activate.  Pour juice into a glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <bowl>", "[GRAB] <bowl>", "[FIND] <food_oatmeal>", "[GRAB] <food_oatmeal>", "[PUTBACK] <food_oatmeal> <bowl>", "[GRAB] <food_oatmeal>", "[PUTOBJBACK] <food_oatmeal>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <bowl>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>", "[FIND] <microwave>", "[OPEN] <microwave>", "[PUTIN] <bowl> <microwave>", "[CLOSE] <microwave>", "[SWITCHON] <microwave>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTOBJBACK] <juice>", "[CLOSE] <freezer>", "[FIND] <table>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'detergent', 'ceilingfan', 'trashcan', 'drawing', 'knife', 'floor', 'ceiling', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'picture', 'couch', 'wall', 'pasta', 'novel', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'maindoor', 'table', 'light', 'vase', 'tooth_paste']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit on my couch and watch TV in my living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'dough', 'drawing', 'wallshelf', 'floor', 'character', 'bedroom', 'purse', 'mat', 'ceilinglamp', 'couch', 'food_jam', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'bed', 'food_food', 'coffe_maker', 'tray', 'towel_rack', 'mouse', 'headset', 'laundry_detergent', 'sink', 'home_office', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Climb into bed, then read the novel", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'freezer', 'detergent', 'floor', 'phone', 'ceiling', 'bedroom', 'keyboard', 'stovefan', 'shower', 'faucet', 'window', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'bench', 'photoframe', 'coffe_maker', 'check', 'television', 'food_food', 'ice', 'rag', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'crayon', 'mop', 'toilet', 'chair', 'cpuscreen', 'floor', 'phone', 'character', 'desk', 'bookshelf', 'stovefan', 'toaster', 'tvstand', 'faucet', 'stereo', 'electric_shaver', 'wall', 'oven', 'food_food', 'television', 'mirror', 'curtain', 'lighter', 'coffee_pot', 'cupboard', 'hairdryer', 'bathroom_cabinet', 'dishwasher']\nTask:Shave\nresponse:\n", "output": {"thoughts": "Pick up razor.  Push the button to turn it on.  Move it in a linear motion against the beard growth.  Make sure to get the bottom of your chin, your cheeks, and mustache area.", "actions": ["[WALK] <bathroom>", "[WALK] <electric_shaver>", "[FIND] <electric_shaver>", "[GRAB] <electric_shaver>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[SWITCHON] <electric_shaver>", "[SWITCHOFF] <electric_shaver>", "[PUTOBJBACK] <electric_shaver>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'detergent', 'trashcan', 'dough', 'clothes_underwear', 'folder', 'stamp', 'bookshelf', 'clothes_jacket', 'table_cloth', 'mousepad', 'couch', 'food_jam', 'clothes_hat', 'pillow', 'nightstand', 'bench', 'oven', 'cutting_board', 'bed', 'food_food', 'check', 'doorjamb', 'console', 'cupboard', 'home_office', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": i go find the phone. i pick up the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'food_carrot', 'milk', 'knifeblock', 'floor', 'ceiling', 'phone', 'bedroom', 'purse', 'toaster', 'stovefan', 'shower', 'faucet', 'couch', 'form', 'pillow', 'wall', 'oven', 'food_food', 'oven_mitts', 'shaving_cream', 'kitchen_counter', 'cd', 'dresser', 'cupboard', 'bathtub', 'table', 'tooth_paste']\nTask:Get glass of milk\nresponse:\n", "output": {"thoughts": "Get an empty glass. Take milk from refrigerator and open it. Pour milk into glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'crayon', 'microwave', 'food_carrot', 'orchid', 'knifeblock', 'bills', 'floor', 'bedroom', 'stamp', 'envelope', 'clothes_jacket', 'bookmark', 'toaster', 'faucet', 'towel', 'dining_room', 'wall', 'coffe_maker', 'food_food', 'video_game_controller', 'tray', 'lighter', 'diary', 'kitchen_counter', 'cupboard', 'sink', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I have had my lunch and walked towards the sink and applied soap to my hands and washed cleanly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['ironing_board', 'sheets', 'tablelamp', 'detergent', 'knifeblock', 'bathroom_counter', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'iron', 'window', 'couch', 'pillow', 'bowl', 'nightstand', 'food_cheese', 'video_game_controller', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'pajamas', 'bag', 'dishwasher', 'board_game', 'clothes_shirt']\nTask:Iron shirt\nresponse:\n", "output": {"thoughts": "Set up an ironing board and an iron.  Lay a shirt on board.  Run iron over shirt until it is free of wrinkles.", "actions": ["[WALK] <bedroom>", "[WALK] <ironing_board>", "[FIND] <ironing_board>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <ironing_board>", "[FIND] <iron>", "[GRAB] <iron>", "[PLUGIN] <iron>", "[PLUGOUT] <iron>", "[PUTOBJBACK] <iron>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'food_butter', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'mop_bucket', 'ceilinglamp', 'tvstand', 'iron', 'window', 'mousepad', 'microphone', 'towel', 'dining_room', 'bowl', 'wall', 'powersocket', 'bench', 'cup', 'novel', 'television', 'coin', 'rag', 'doorjamb', 'headset', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk in to the living room. I find the light switch. I turn on the lights.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'knifeblock', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'tea_bag', 'computer', 'deck_of_cards', 'mousepad', 'couch', 'pillow', 'wall', 'child', 'bed', 'coin', 'coffee_filter', 'doorjamb', 'centerpiece', 'door', 'table', 'light', 'dishwasher']\nTask:Let baby learn how to walk\nresponse:\n", "output": {"thoughts": "The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[GREET] <child>", "[FIND] <couch>", "[TURNTO] <couch>", "[POINTAT] <couch>", "[TURNTO] <child>", "[LOOKAT] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blow_dryer', 'bathroom', 'food_carrot', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stamp', 'bedroom', 'food_cake', 'computer', 'shower', 'picture', 'oven', 'basket_for_clothes', 'photoframe', 'food_food', 'check', 'towel_rack', 'mouse', 'pencil', 'shoes', 'comb', 'hairbrush', 'cupboard', 'laundry_detergent', 'sink', 'napkin', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Pick up my dirty clothes, load the washer, add detergent, turn on washer,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[RUN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'sponge', 'tablelamp', 'mop', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'stamp', 'wine_glass', 'clothes_socks', 'couch', 'pillow', 'wall', 'nightstand', 'food_cheese', 'bed', 'photoframe', 'food_food', 'check', 'doorjamb', 'door', 'sink', 'bag', 'bathroom_cabinet', 'table', 'light']\nTask:Entertain\nresponse:\n", "output": {"thoughts": "Give guests wine, cheese, crackers and sit and talk", "actions": ["[WALK] <home_office>", "[WALK] <food_cheese>", "[FIND] <food_cheese>", "[FIND] <food_food>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <wine_glass>", "[GRAB] <wine_glass>", "[DRINK] <wine_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'walllamp', 'orchid', 'floor', 'ceiling', 'character', 'desk', 'table_cloth', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'stovefan', 'dish_soap', 'mousepad', 'couch', 'wall', 'powersocket', 'bed', 'coffe_maker', 'food_food', 'kitchen_counter', 'cupboard', 'controller', 'sink', 'dishwasher', 'soap', 'clothes_shirt', 'teeth']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'walllamp', 'nail_polish', 'knifeblock', 'drawing', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'character', 'desk', 'faucet', 'stereo', 'shower', 'mousepad', 'couch', 'wall', 'oven', 'cup', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'chessboard', 'cupboard', 'sink', 'table', 'dirt']\nTask:Get glass of water\nresponse:\n", "output": {"thoughts": "Get a glass from a cabinet. Hold it under the sink and turn the cold water on. Fill the glass. Turn off the water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <kitchen_cabinet>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'measuring_cup', 'ground_coffee', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'keyboard', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'oven', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'blender', 'chair', 'freezer', 'walllamp', 'nail_polish', 'detergent', 'drawing', 'knife', 'wallshelf', 'bathroom_counter', 'floor', 'desk', 'computer', 'mat', 'ceilinglamp', 'plate', 'faucet', 'mousepad', 'candle', 'microphone', 'bowl', 'wall', 'nightstand', 'bench', 'bed', 'keys', 'food_food', 'table']\nTask:Eat\nresponse:\n", "output": {"thoughts": "I get my prepared dinner set on the dining table, sit down on chair, pick up utensils, eat food", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <food_food> <table>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <plate>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'tea', 'band-aids', 'walllamp', 'nail_polish', 'knifeblock', 'bills', 'cpuscreen', 'hanger', 'shampoo', 'phone', 'floor', 'bedroom', 'desk', 'toaster', 'stereo', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bowl', 'dining_room', 'bench', 'newspaper', 'clothes_gloves', 'coffe_maker', 'curtain', 'diary', 'hairbrush', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'microwave', 'walllamp', 'drawing', 'knifeblock', 'cpuscreen', 'orchid', 'wallshelf', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'faucet', 'wall', 'oven', 'powersocket', 'bench', 'after_shave', 'curtain', 'centerpiece', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'soap']\nTask:Wipe down counter\nresponse:\n", "output": {"thoughts": "Clean counter", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <rag>", "[SQUEEZE] <rag>", "[SWITCHOFF] <faucet>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": you go into your office, switch on the light, find chair at desk, sit in the chair, switch on the computer and work.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['scissors', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'window', 'toaster', 'clothes_pants', 'broom', 'powersocket', 'bench', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'dresser', 'hair', 'door', 'sink', 'clothes_shirt']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "Get some clothes together that make an appropriate outfit.  Put them on", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'stovefan', 'toaster', 'laptop', 'faucet', 'closetdrawer', 'couch', 'bowl', 'dining_room', 'wall', 'oven', 'bench', 'bed', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I switch on the stove and cook my groceries there.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <oven>', '[SWITCHON] <oven>', '[FIND] <bowl>', '[GRAB] <bowl>', '[PUTBACK] <bowl> <oven>', '[FIND] <food_food>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'ceiling', 'phone', 'desk', 'bookshelf', 'cleaning_solution', 'mat', 'tvstand', 'faucet', 'shower', 'couch', 'wall', 'nightstand', 'bench', 'toy', 'photoframe', 'mirror', 'tray', 'towel_rack', 'paper_towel', 'door', 'sink', 'clothes_scarf', 'bathroom_cabinet']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[CLOSE] <bathroom_cabinet>", "[WALK] <mirror>", "[FIND] <cleaning_solution>", "[RINSE] <cleaning_solution>", "[WIPE] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'microwave', 'walllamp', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'bedroom', 'ceilinglamp', 'plate', 'faucet', 'window', 'dish_soap', 'mousepad', 'bowl', 'wall', 'oven', 'powersocket', 'check', 'mouthwash', 'after_shave', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": I go to the kitchen and turn on the faucet. I get dish soap and put it into the sink. I grab a sponge and a bowl and start cleaning it. I put the bowl into the dishrack. I find a plate and clean it as well. I put it into the dishrack, turn off the faucet and put the sponge back.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <bowl>', '[GRAB] <bowl>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <bowl>', '[SCRUB] <bowl>', '[RINSE] <bowl>', '[FIND] <dishrack>', '[PUTBACK] <bowl> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <sponge>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'folder', 'tea_bag', 'mat', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'pasta', 'nightstand', 'photoframe', 'novel', 'food_food', 'tray', 'curtain', 'doorjamb', 'door', 'table', 'light', 'board_game']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Bring book, lie on the sofa, read book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[LIE] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'trashcan', 'orchid', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'bookshelf', 'clothes_jacket', 'faucet', 'couch', 'clothes_pants', 'form', 'wall', 'cutting_board', 'photoframe', 'check', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'clothes_dress', 'door', 'sink', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open cabinet and collect clothes and put them into washing machine and wash clothes,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_jacket>', '[GRAB] <clothes_jacket>', '[PUTBACK] <clothes_jacket> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'drawing', 'ceiling', 'dvd_player', 'character', 'bookshelf', 'keyboard', 'brush', 'ceilinglamp', 'toaster', 'mat', 'window', 'nightstand', 'bench', 'photoframe', 'food_food', 'after_shave', 'diary', 'kettle', 'shoes', 'chessboard', 'dresser', 'door', 'bathroom_cabinet', 'colander']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Using soft brush, carefully dust both Indonesian wedding doll statues.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <brush>", "[GRAB] <brush>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'trashcan', 'food_carrot', 'orchid', 'floor', 'ceiling', 'thread', 'mail', 'button', 'food_bacon', 'bookmark', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'wall', 'bench', 'cup', 'food_food', 'razor', 'television', 'food_donut', 'towel_rack', 'console', 'kitchen_counter', 'hairbrush', 'cupboard', 'sink', 'vase']\nTask:Added meat to freezer\nresponse:\n{\n    \"thoughts\": After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['microwave', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'stovefan', 'mat', 'window', 'electric_shaver', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'oven', 'toy', 'bed', 'mouthwash', 'tray', 'doorjamb', 'mouse', 'pencil', 'chessboard', 'door', 'bathtub', 'instrument_guitar', 'light', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Hit light switch", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'tablelamp', 'blender', 'scissors', 'freezer', 'walllamp', 'ground_coffee', 'orchid', 'character', 'bedroom', 'bookshelf', 'alcohol', 'button', 'toaster', 'stereo', 'homework', 'electric_shaver', 'mousepad', 'couch', 'bowl', 'toy', 'clothes_gloves', 'coffe_maker', 'kitchen_counter', 'conditioner', 'napkin', 'table', 'light', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room, look for radio, grab on button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'toilet', 'chair', 'bathroom', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'computer', 'tvstand', 'faucet', 'stereo', 'food_orange', 'photoframe', 'novel', 'television', 'curtain', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'shoes', 'dresser', 'cupboard', 'sink', 'bathtub', 'home_office']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "I get dressed in my bedroom in the mornings.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'tea_bag', 'mat', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'toy', 'photoframe', 'check', 'mouthwash', 'food_food', 'food_donut', 'wine', 'curtain', 'doorjamb', 'console', 'dresser', 'door', 'light']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'measuring_cup', 'microwave', 'chair', 'freezer', 'band-aids', 'walllamp', 'detergent', 'knifeblock', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'mop_bucket', 'toaster', 'stovefan', 'faucet', 'couch', 'wall', 'oven', 'cup', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'food_snack', 'sink', 'napkin', 'home_office']\nTask:Have snack\nresponse:\n", "output": {"thoughts": "Pour soda into a glass. Get pretzels out of bag. Eat and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTIN] <juice> <freezer>", "[CLOSE] <freezer>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <food_snack>", "[GRAB] <food_snack>", "[CLOSE] <cupboard>", "[DRINK] <cup>", "[EAT] <food_snack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'pot', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'oil', 'mousepad', 'food_peanut_butter', 'wall', 'oven', 'powersocket', 'food_food', 'water', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sheets', 'pot', 'tablelamp', 'tea', 'microwave', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'knifeblock', 'bathroom_counter', 'floor', 'remote_control', 'table_cloth', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'tvstand', 'wall', 'nightstand', 'bed', 'newspaper', 'photoframe', 'television', 'food_rice', 'console', 'pencil', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watching tv with a remote is always fun; lay back, relax and enjoy;", "actions": ["[WALK] <bedroom>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'doorjamb', 'towel_rack', 'pencil', 'shoes', 'dresser', 'sink', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go to the the living room, sit down on the sofa, and then I switch on the TV with the TV remote,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'chair', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'stereo', 'dining_room', 'wall', 'nightstand', 'powersocket', 'bed', 'photoframe', 'video_game_controller', 'curtain', 'doorjamb', 'lighter', 'kitchen_counter', 'dresser', 'cupboard', 'bathroom_cabinet', 'instrument_guitar']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Put iPhone into speaker holder, open up the Pandora app, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <stereo>", "[PLUGIN] <phone>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'detergent', 'cat', 'drawing', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'dustpan', 'pillow', 'wall', 'wall_clock', 'television', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'table', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I see my cat on the couch so I walk over, sit down and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'bathroom', 'detergent', 'wallshelf', 'hanger', 'floor', 'character', 'keyboard', 'envelope', 'clothes_jacket', 'stovefan', 'stereo', 'window', 'food_peanut_butter', 'nightstand', 'clothes_gloves', 'food_food', 'food_donut', 'after_shave', 'centerpiece', 'towel_rack', 'cupboard', 'door', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'table', 'home_office']\nTask:Put on coat\nresponse:\n", "output": {"thoughts": "I remove my scarf from a coat peg in the entrance hall.  I wrap the scarf around my neck", "actions": ["[WALK] <home_office>", "[WALK] <hanger>", "[TURNTO] <hanger>", "[LOOKAT] <hanger>", "[FIND] <clothes_scarf>", "[GRAB] <clothes_scarf>", "[PUTON] <clothes_scarf>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'toilet', 'freezer', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'ceiling', 'phone', 'character', 'keyboard', 'toaster', 'mat', 'floor_lamp', 'oil', 'pillow', 'dining_room', 'nightstand', 'wall_clock', 'bench', 'novel', 'television', 'curtain', 'doorjamb', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Turn a light on and read a book in the bedroom.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'beer', 'microwave', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'dough', 'orchid', 'dvd_player', 'keyboard', 'piano_bench', 'brush', 'faucet', 'pillow', 'dining_room', 'powersocket', 'newspaper', 'coffe_maker', 'television', 'wine', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'conditioner', 'clothes_scarf', 'bathtub', 'bathroom_cabinet', 'creditcard']\nTask:Lock door\nresponse:\n", "output": {"thoughts": "Turn the lock.", "actions": ["[WALK] <bedroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'freezer', 'walllamp', 'trashcan', 'food_carrot', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bedroom', 'character', 'ceilinglamp', 'tvstand', 'faucet', 'dustpan', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'photoframe', 'bed', 'food_food', 'television', 'food_donut', 'food_rice', 'doorjamb', 'shaving_cream', 'table', 'vase']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": walk to kitchen and seprate the groceries and arrange in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'centerpiece', 'kitchen_counter', 'door', 'table', 'light', 'creditcard']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I like to set on my sofa.Turn on the tv. And watch Perry Mason. Monday's through Friday's.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'food_salt', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'purse', 'computer', 'ceilinglamp', 'window', 'mousepad', 'food_noodles', 'wall', 'powersocket', 'clothes_gloves', 'food_food', 'toilet_paper', 'television', 'check', 'tray', 'doorjamb', 'cards', 'mouse', 'door', 'light', 'table']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Open the fridge, put the groceries in the fridge and close it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'walllamp', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch. Get comfortable. Switch on the television. Stare at the television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'computer', 'mat', 'shoe-shine_kit', 'oil', 'pillow', 'wall', 'pasta', 'powersocket', 'cup', 'food_food', 'video_game_controller', 'toothbrush_holder', 'television', 'doorjamb', 'kettle', 'door', 'light', 'clothes_skirt', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on light in an area,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['beer', 'microwave', 'toilet', 'chair', 'freezer', 'wallshelf', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'toaster', 'ceilinglamp', 'window', 'pillow', 'wall', 'powersocket', 'oven', 'bench', 'food_cheese', 'bed', 'photoframe', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'lighter', 'kitchen_counter', 'cupboard', 'home_office']\nTask:Eat cheese\nresponse:\n", "output": {"thoughts": "Open fridge.  Get cheese.  Eat it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[EAT] <food_cheese>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'window', 'clothes_pants', 'pillow', 'wall', 'wall_clock', 'oven', 'nightstand', 'cutting_board', 'shelf', 'clothes_gloves', 'food_food', 'coffe_maker', 'wooden_spoon', 'tray', 'kitchen_counter', 'sink', 'bathroom_cabinet']\nTask:Keep an eye on stove as something is cooking\nresponse:\n{\n    \"thoughts\": Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <pot>', '[TURNTO] <pot>', '[LOOKAT] <pot>', '[FIND] <wall_clock>', '[TURNTO] <wall_clock>', '[LOOKAT] <wall_clock>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['band-aids', 'trashcan', 'woman', 'drawing', 'orchid', 'floor', 'ceiling', 'man', 'envelope', 'tvstand', 'stereo', 'shoe-shine_kit', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'cup', 'food_food', 'coffe_maker', 'razor', 'tray', 'doorjamb', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table', 'light', 'home_office']\nTask:Hang with friends\nresponse:\n", "output": {"thoughts": "Sit with friends. Talk", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'bookshelf', 'keyboard', 'brush', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'stovefan', 'shoe-shine_kit', 'pillow', 'wall', 'powersocket', 'photoframe', 'novel', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'bathroom_cabinet', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit on the couch. Turn on TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'detergent', 'drawing', 'knifeblock', 'food_egg', 'floor', 'computer', 'toaster', 'stovefan', 'tvstand', 'stereo', 'window', 'food_noodles', 'mousepad', 'couch', 'pillow', 'dining_room', 'nightstand', 'pasta', 'bed', 'oven_mitts', 'tray', 'curtain', 'lighter', 'towel_rack', 'shoes', 'cd', 'pajamas']\nTask:Put clothes away\nresponse:\n", "output": {"thoughts": "I take clean clothes and either place them in dresser drawers or hang them in a closet.", "actions": ["[WALK] <bedroom>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'microwave', 'toilet', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'mop_bucket', 'stovefan', 'mat', 'faucet', 'mousepad', 'towel', 'wall', 'wall_clock', 'oven', 'bench', 'cutting_board', 'shelf', 'food_rice', 'tray', 'after_shave', 'towel_rack', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'soap', 'laser_pointer', 'dirt', 'hands_both']\nTask:Hand washing\nresponse:\n{\n    \"thoughts\": Turn on the faucet. Soap up your hands. Then rinse under the running water of the faucet. Turn off faucet and dry hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[WASH] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'ground_coffee', 'walllamp', 'trashcan', 'knifeblock', 'hanger', 'floor', 'phone', 'character', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'mousepad', 'pillow', 'wall', 'oven', 'cutting_board', 'television', 'food_rice', 'water', 'tray', 'kitchen_counter', 'cd', 'dresser', 'cupboard', 'bathroom_cabinet', 'home_office']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Get my coffee pot going", "actions": ["[WALK] <dining_room>", "[WALK] <coffee_pot>", "[FIND] <coffee_pot>", "[GRAB] <coffee_pot>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffee_pot>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffee_pot>", "[GRAB] <ground_coffee>", "[PUTOBJBACK] <ground_coffee>", "[WALK] <oven>", "[PUTBACK] <coffee_pot> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'food_ice_cream', 'hanger', 'floor', 'ceiling', 'character', 'bookshelf', 'mat', 'tvstand', 'faucet', 'ceilinglamp', 'wall', 'powersocket', 'photoframe', 'clothes_gloves', 'bed', 'food_food', 'check', 'ice', 'doorjamb', 'pencil', 'kitchen_counter', 'laundry_detergent', 'door', 'napkin', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on the light,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'woman', 'knifeblock', 'needle', 'cpuscreen', 'desk', 'man', 'keyboard', 'bedroom', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'nightstand', 'oven', 'cup', 'food_food', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'food_bacon', 'toaster', 'tvstand', 'window', 'oil', 'couch', 'wall', 'nightstand', 'newspaper', 'coffe_maker', 'television', 'coffee_filter', 'water', 'food_donut', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'dirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'man', 'bookshelf', 'clothes_jacket', 'piano_bench', 'toaster', 'stereo', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'coffe_maker', 'doorjamb', 'cupboard', 'door', 'sink', 'home_office', 'table', 'light']\nTask:Have conversation with boyfriend\nresponse:\n", "output": {"thoughts": "Sit on sofa, talk", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'tablelamp', 'microwave', 'hanger', 'floor', 'ceiling', 'toaster', 'mat', 'faucet', 'shower', 'mousepad', 'couch', 'candle', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'drinking_glass', 'wooden_spoon', 'curtain', 'tray', 'doorjamb', 'kitchen_counter', 'paper_towel', 'cupboard', 'sink', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": wipe loose dirt, place in the dishwasher, turn on the dishwasher,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[WIPE] <drinking_glass>', '[RINSE] <drinking_glass>', '[PUTBACK] <drinking_glass> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'mop', 'microwave', 'walllamp', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'iron', 'window', 'laptop', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'check', 'tray', 'rag', 'towel_rack', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet']\nTask:Wash dishes\nresponse:\n", "output": {"thoughts": "Run hot water into tub of dirty dishes in the sink.  Add liquid soap.  Scrub dishes with wash rag. Rinse and place on drying rack.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sink>", "[PUTOBJBACK] <dish_soap>", "[FIND] <rag>", "[GRAB] <rag>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'chair', 'bathroom', 'drawing', 'wallshelf', 'cpuscreen', 'character', 'desk', 'envelope', 'computer', 'shower', 'faucet', 'electric_shaver', 'dustpan', 'dining_room', 'wall', 'bench', 'tape', 'photoframe', 'check', 'doorjamb', 'centerpiece', 'mouse', 'hairbrush', 'bathroom_cabinet', 'table', 'home_office', 'light', 'instrument_guitar', 'creditcard']\nTask:Receive credit card\nresponse:\n{\n    \"thoughts\": hear the dooble, go to door, take de envelope, open it with a scissor, take the creditcard,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <envelope>', '[FIND] <envelope>', '[GRAB] <envelope>', '[WALK] <home_office>', '[WALK] <scissors>', '[FIND] <scissors>', '[OPEN] <envelope>', '[FIND] <creditcard>', '[GRAB] <creditcard>']\n}\n", "input": "\nScene_Objects:['microwave', 'cat', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'mousepad', 'bench', 'newspaper', 'coffe_maker', 'food_food', 'tray', 'curtain', 'mouse', 'door', 'hairdryer', 'sink', 'food_steak', 'light', 'dishwasher', 'spoon', 'table', 'board_game']\nTask:Write  school paper\nresponse:\n", "output": {"thoughts": "Writing  and drafting a paper for  class work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'microwave', 'freezer', 'drawing', 'phone', 'bookshelf', 'toaster', 'stovefan', 'tvstand', 'mousepad', 'couch', 'clothes_hat', 'food_orange', 'wall', 'photoframe', 'clothes_gloves', 'bed', 'check', 'television', 'food_rice', 'tray', 'doorjamb', 'dresser', 'hairdryer', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'door']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to my bedroom, and then I switch on the radio,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['scissors', 'ground_coffee', 'trashcan', 'woman', 'knife', 'cpuscreen', 'phone', 'man', 'bookshelf', 'computer', 'stovefan', 'mat', 'brush', 'shower', 'deck_of_cards', 'pillow', 'powersocket', 'bench', 'bed', 'coffe_maker', 'television', 'wooden_spoon', 'doorjamb', 'mouse', 'kitchen_counter', 'chessboard', 'door', 'bathtub', 'bathroom_cabinet', 'coffee']\nTask:Greet people\nresponse:\n", "output": {"thoughts": "Say HI to those that come in.", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'powersocket', 'clothes_gloves', 'television', 'razor', 'curtain', 'doorjamb', 'console', 'mouse', 'door', 'bag', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk in living room. I find TV remote. I grab TV remote. I sit on sofa. I switch on TV. I switch channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'dresser', 'door', 'home_office', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch nd watch t.V", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'ceilinglamp', 'shoe-shine_kit', 'faucet', 'clothes_pants', 'dining_room', 'wall', 'tape', 'basket_for_clothes', 'bed', 'coffe_maker', 'chessboard', 'clothes_dress', 'door', 'home_office', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'bathroom', 'detergent', 'cat', 'trashcan', 'woman', 'drawing', 'character', 'bedroom', 'man', 'keyboard', 'bookshelf', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'clothes_hat', 'dining_room', 'wall', 'child', 'toy', 'tray', 'curtain', 'mouse', 'cupboard', 'door', 'spoon', 'home_office', 'table']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet guests when they knock on the door and invite them to enter my home", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'microwave', 'ceilingfan', 'cpuscreen', 'hanger', 'bathroom_counter', 'character', 'keyboard', 'toaster', 'mat', 'balanceball', 'stereo', 'window', 'microphone', 'bench', 'basket_for_clothes', 'clothes_gloves', 'check', 'oven_mitts', 'kettle', 'dresser', 'cupboard', 'maindoor', 'napkin', 'light', 'clothes_skirt', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Wash the clothes in the washing machine,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'table_cloth', 'faucet', 'homework', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'photoframe', 'bed', 'doorjamb', 'cards', 'cupboard', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Arrange furniture\nresponse:\n", "output": {"thoughts": "Move furniture where you want it.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[PUSH] <coffee_table>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'cutting_board', 'cup', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'controller', 'door', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'jelly', 'chair', 'walllamp', 'detergent', 'trashcan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'keyboard', 'stovefan', 'mat', 'ceilinglamp', 'food_orange', 'wall', 'bench', 'oven', 'food_food', 'oven_mitts', 'doorjamb', 'dresser', 'cupboard', 'home_office', 'table', 'bathroom_cabinet']\nTask:Play games\nresponse:\n", "output": {"thoughts": "I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.", "actions": ["[WALK] <home_office>", "[WALK] <board_game>", "[FIND] <board_game>", "[GRAB] <board_game>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <board_game> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'microwave', 'chair', 'band-aids', 'woman', 'needle', 'longboard', 'man', 'keyboard', 'envelope', 'toaster', 'mat', 'ceilinglamp', 'shower', 'faucet', 'stereo', 'nightstand', 'box', 'light_bulb', 'coffe_maker', 'curtain', 'towel_rack', 'shoes', 'dresser', 'hairbrush', 'door', 'vase', 'instrument_violin', 'creditcard']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'food_butter', 'knifeblock', 'floor', 'ceiling', 'character', 'desk', 'food_cake', 'computer', 'bookmark', 'toaster', 'stovefan', 'mat', 'faucet', 'food_kiwi', 'wall', 'oven', 'powersocket', 'food_vegetable', 'bed', 'food_food', 'oven_mitts', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I prepare meals and cook them on the stove.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[FIND] <oven>", "[SWITCHON] <oven>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'chair', 'walllamp', 'blow_dryer', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'ceiling', 'thread', 'ceilinglamp', 'laptop', 'faucet', 'mousepad', 'couch', 'candle', 'food_jam', 'pillow', 'wall', 'nightstand', 'tape', 'television', 'tray', 'after_shave', 'food_onion', 'shoes', 'cupboard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I sit in the sofa and I relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['blender', 'juice', 'cd_player', 'chair', 'freezer', 'detergent', 'trashcan', 'knife', 'knifeblock', 'drawing', 'floor', 'ceiling', 'bedroom', 'keyboard', 'remote_control', 'computer', 'button', 'ceilinglamp', 'faucet', 'mousepad', 'wall', 'nightstand', 'after_shave', 'door', 'bathtub', 'home_office', 'table', 'clothes_skirt', 'instrument_guitar', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.", "actions": ["[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'mop', 'walllamp', 'detergent', 'food_carrot', 'toothbrush', 'knifeblock', 'orchid', 'clothes_underwear', 'hanger', 'phone', 'stovefan', 'shower', 'homework', 'couch', 'food_jam', 'microphone', 'bowl', 'photoframe', 'check', 'curtain', 'console', 'mouse', 'pencil', 'dresser', 'sink', 'soap', 'creditcard']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'toilet', 'detergent', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'bedroom', 'folder', 'mat', 'shower', 'food_noodles', 'couch', 'pillow', 'wall', 'oven', 'food_food', 'coffe_maker', 'doorjamb', 'door', 'bathtub', 'home_office', 'table', 'light', 'clothes_skirt', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'toilet', 'bathroom', 'detergent', 'ceilingfan', 'food_butter', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'pillow', 'wall', 'food_food', 'television', 'tray', 'curtain', 'after_shave', 'cupboard', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": TURN ON TV, SIT ON COUCH ACROSS THE ROOM, USE REMOTE TO LOOK FOR SHOW TO WATCH,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'walllamp', 'knifeblock', 'food_egg', 'kitchen_cabinet', 'floor', 'phone', 'ceiling', 'food_chicken', 'stovefan', 'toaster', 'faucet', 'form', 'wall', 'powersocket', 'oven', 'bench', 'coffe_maker', 'tray', 'doorjamb', 'lighter', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'food_steak', 'sink', 'light']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and walk over the cabinent. I pull out a pan and walk to the stove. I put the pan on the stove and walk over to the fridge. I pull out out meat from the fridge and walk back over the stove. Then I put the meat in the pan.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[WALK] <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'desk', 'envelope', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'oven_mitts', 'curtain', 'controller', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room. I cross the room to the couch. I sit down and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'dvd_player', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I go in my living room, find the remote, sit on my couch,  turn on the TV and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'chair', 'nail_polish', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'ceilinglamp', 'balanceball', 'shower', 'mousepad', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'toothbrush_holder', 'doorjamb', 'mouse', 'light', 'vase']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['chair', 'bathroom', 'orchid', 'wallshelf', 'ceiling', 'phone', 'stamp', 'bookshelf', 'mail', 'computer', 'tvstand', 'shower', 'couch', 'pillow', 'oven', 'powersocket', 'box', 'bed', 'clothes_gloves', 'photoframe', 'check', 'curtain', 'pencil', 'dresser', 'cupboard', 'door', 'sink', 'bathtub', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find a cat, touch it gently", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'beer', 'measuring_cup', 'toilet', 'freezer', 'food_carrot', 'drawing', 'food_egg', 'hanger', 'floor', 'ceiling', 'bedroom', 'mat', 'laptop', 'shower', 'tvstand', 'closetdrawer', 'food_orange', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'drinking_glass', 'wine', 'doorjamb', 'door', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Robot Drinks the water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>', '[PUTOBJBACK] <drinking_glass>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'walllamp', 'drawing', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'stovefan', 'plate', 'faucet', 'tvstand', 'window', 'dish_soap', 'towel', 'dining_room', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'food_food', 'curtain', 'kitchen_counter', 'cupboard', 'food_steak', 'sink']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walk to Kitchen, find sink, find dishes and dish soap. Put soap on plate. open faucet, rinse dishes close faucet.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <plate>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'freezer', 'knifeblock', 'drawing', 'wallshelf', 'ceiling', 'phone', 'desk', 'keyboard', 'face', 'toaster', 'laptop', 'tvstand', 'shower', 'stereo', 'homework', 'oil', 'towel', 'oven', 'basket_for_clothes', 'tray', 'towel_rack', 'hair', 'cupboard', 'door', 'bathtub', 'clothes_shirt']\nTask:Rain welcome\nresponse:\n{\n    \"thoughts\": The robot would greet me and then hand me a towel.  It would make an appropriate quip about the weather while I dried off.  It would receive the wet towel when I was done.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <towel>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <face>', '[WIPE] <face>', '[FIND] <hair>', '[WIPE] <hair>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'picture', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <chair>", "[SIT] <chair>", "[TURNTO] <television>", "[WATCH] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'chair', 'bathroom', 'detergent', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'mat', 'tvstand', 'clothes_pants', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'novel', 'check', 'keys', 'tray', 'after_shave', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'tooth_paste']\nTask:Read\nresponse:\n{\n    \"thoughts\": I like to read books in my bedroom.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <desk>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <bed>', '[LIE] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['blow_dryer', 'walllamp', 'bathroom_counter', 'phone', 'desk', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'button', 'clothes_jacket', 'stovefan', 'toaster', 'window', 'couch', 'bench', 'oven', 'cup', 'coffe_maker', 'television', 'ice', 'towel_rack', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'bathtub', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'mop', 'walllamp', 'trashcan', 'wallshelf', 'floor', 'ceiling', 'folder', 'bedroom', 'tvstand', 'shower', 'foundation', 'window', 'mousepad', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'food_food', 'coffe_maker', 'television', 'food_donut', 'curtain', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to the bathroom and wash with the shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>', '[WASH] <shower>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'jelly', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'cleaning_solution', 'toaster', 'mat', 'faucet', 'window', 'couch', 'bowl', 'wall', 'oven', 'cup', 'food_food', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'instrument_violin', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Washing hands", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <cleaning_solution>", "[SQUEEZE] <cleaning_solution>", "[SCRUB] <hands_both>", "[SWITCHON] <faucet>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'ground_coffee', 'walllamp', 'detergent', 'knifeblock', 'longboard', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'faucet', 'candle', 'wall', 'box', 'coffe_maker', 'food_food', 'television', 'coffee_filter', 'water', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'hair', 'maindoor', 'sink', 'napkin', 'dishwasher', 'tooth_paste']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'trashcan', 'desk', 'keyboard', 'stovefan', 'ceilinglamp', 'mat', 'toaster', 'stereo', 'window', 'couch', 'dining_room', 'nightstand', 'wall', 'cutting_board', 'bed', 'photoframe', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'hairdryer', 'bathtub', 'bathroom_cabinet', 'home_office', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into room, grab phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'food_pizza', 'microwave', 'dog', 'band-aids', 'orchid', 'wallshelf', 'floor', 'desk', 'mat', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'tape', 'cutting_board', 'bed', 'standingmirror', 'television', 'curtain', 'doorjamb', 'conditioner', 'sink', 'home_office', 'table', 'light', 'bathroom_cabinet', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'freezer', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'orchid', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'ceilinglamp', 'tvstand', 'faucet', 'wall', 'nightstand', 'curtain', 'doorjamb', 'chessboard', 'shoes', 'cupboard', 'sink', 'bathroom_cabinet', 'home_office', 'dishwasher', 'light', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash Hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <faucet>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'orchid', 'needle', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to my desk. I turn on the computer. I open a web page and look at it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['toilet', 'band-aids', 'drawing', 'orchid', 'kitchen_cabinet', 'floor', 'bathroom_counter', 'phone', 'bookshelf', 'computer', 'stovefan', 'mat', 'plate', 'ceilinglamp', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'bench', 'powersocket', 'oven', 'cutting_board', 'curtain', 'mouse', 'table', 'dishwasher', 'light']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the kitchen cabinet. Grab the plate and put it on the table.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'faucet', 'picture', 'window', 'mousepad', 'oil', 'clothes_hat', 'wall', 'keys', 'food_food', 'coin', 'food_onion', 'mouse', 'cupboard', 'laundry_detergent', 'door', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I walk into the home office and walk to the desk. I pull out the chair and sit down. I turn on the computer. After the computer is turned on I click on my email program.  When the program is loaded I write an email.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'toilet', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'plate', 'faucet', 'drying_rack', 'dish_soap', 'couch', 'dining_room', 'wall', 'oven', 'newspaper', 'photoframe', 'check', 'toilet_paper', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'table']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I walk in to the Kitchen. I find a plate, a sponge, and the dish soap. I find the faucet. I turn the faucet on. I pour water on the sponge and plate. I add dish soap to the sponge. I put the dish soap back. I scrub the plate. I rinse the plate. I place the plate on the drying rack. I rinse the sponge. I turn the faucet off.", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <sponge>", "[GRAB] <sponge>", "[RINSE] <sponge>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>", "[RINSE] <plate>", "[FIND] <drying_rack>", "[PUTBACK] <plate> <plate>", "[RINSE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'microwave', 'bathroom', 'trashcan', 'toothbrush', 'drawing', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'purse', 'food_bacon', 'ceilinglamp', 'plate', 'laptop', 'deck_of_cards', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'spoon', 'home_office', 'table', 'teeth']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'tablelamp', 'microwave', 'drawing', 'orchid', 'kitchen_cabinet', 'floor', 'phone', 'bookshelf', 'computer', 'stovefan', 'mat', 'plate', 'ceilinglamp', 'deck_of_cards', 'faucet', 'couch', 'wall', 'nightstand', 'bench', 'oven', 'box', 'cup', 'check', 'razor', 'food_onion', 'dresser', 'door', 'table']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Grab plates and forks. Walk into dining room. Put plates and forks on table.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <fork> <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ceilingfan', 'toothbrush', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'longboard', 'mop_bucket', 'mat', 'tvstand', 'faucet', 'couch', 'clothes_pants', 'towel', 'wall', 'basket_for_clothes', 'photoframe', 'light_bulb', 'wooden_spoon', 'after_shave', 'coffee', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'laser_pointer', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Turn faucet clockwise to turn on water.  Place hands under stream of water, pick up soap, rub hands with soap.  Put soap back down.  Rub hands together with soap on them.  Rinse hands under water from faucet.  Turn faucet counterclockwise to turn water off.  Dry hands by rubbing them on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[WASH] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn on the TV, sit on the couch, find a show, and watch it.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'chair', 'jelly', 'knifeblock', 'floor', 'ceiling', 'character', 'alcohol', 'toaster', 'plate', 'faucet', 'drying_rack', 'window', 'tvstand', 'dish_soap', 'closetdrawer', 'couch', 'towel', 'pillow', 'food_kiwi', 'wall', 'toy', 'novel', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'instrument_guitar', 'shoe_rack']\nTask:Do dishes\nresponse:\n{\n    \"thoughts\": Wash dirty dishes, dry them and then put them away,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[FIND] <drying_rack>', '[PUTBACK] <plate> <drying_rack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <drying_rack>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'chair', 'freezer', 'detergent', 'bathroom_counter', 'desk', 'bedroom', 'keyboard', 'button', 'stovefan', 'ceilinglamp', 'stereo', 'shower', 'window', 'clothes_socks', 'dining_room', 'nightstand', 'wall', 'bench', 'oven', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'dresser', 'napkin', 'bathroom_cabinet', 'soap']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room, look for radio, grab on button", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'bathtub', 'sink', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room and go to sofa and turn on tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[WALK] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[DROP] <remote_control>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'ground_coffee', 'walllamp', 'bathroom', 'knifeblock', 'drawing', 'floor', 'ceiling', 'character', 'bookshelf', 'toaster', 'stovefan', 'faucet', 'wall', 'oven', 'tape', 'cup', 'coffe_maker', 'food_food', 'toilet_paper', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'table', 'dishwasher', 'shoe_rack']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'food_butter', 'food_carrot', 'drawing', 'orchid', 'knifeblock', 'knife', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'toaster', 'mat', 'food_noodles', 'couch', 'pillow', 'wall', 'bed', 'photoframe', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'cupboard', 'door', 'bathroom_cabinet', 'table', 'light', 'dirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Relax on the sofa in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'food_pizza', 'walllamp', 'orchid', 'needle', 'cpuscreen', 'bathroom_counter', 'shampoo', 'desk', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'shower', 'faucet', 'couch', 'dining_room', 'cup', 'bed', 'tray', 'curtain', 'wooden_spoon', 'pencil', 'door', 'bathroom_cabinet', 'light', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office. I find the telephone. I pick up the telephone handset.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'jelly', 'walllamp', 'bathroom', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'couch', 'food_jam', 'wall', 'novel', 'television', 'curtain', 'doorjamb', 'comb', 'clothes_dress', 'sink', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'faucet', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'cupboard', 'door', 'home_office', 'light', 'table', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room. Find remote control. Look at remote control. Push power on button. Find sofa. Sit down on sofa. Watch Television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'button', 'food_bacon', 'stovefan', 'mat', 'shower', 'faucet', 'closetdrawer', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'keys', 'tray', 'doorjamb', 'lighter', 'sink', 'bathtub', 'bathroom_cabinet', 'soap']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": I want to take a shower so I go to the bathroom, turn on the shower, get the soap, lather and wash, then I rinse, turn off shower, get a towel and dry off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[RINSE] <shower>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>']\n}\n", "input": "\nScene_Objects:['walllamp', 'cat', 'trashcan', 'drawing', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bookshelf', 'keyboard', 'mail', 'toaster', 'mat', 'tvstand', 'couch', 'pillow', 'broom', 'wall', 'nightstand', 'powersocket', 'television', 'food_donut', 'doorjamb', 'door', 'bathroom_cabinet', 'light', 'table', 'dirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "My cat looks like it wants attention, so I will go pet it on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'juice', 'toilet', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'food_peanut_butter', 'food_jam', 'bowl', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'food_steak', 'light', 'dishwasher', 'table']\nTask:Grab some juice\nresponse:\n{\n    \"thoughts\": I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['fork', 'food_bread', 'tablelamp', 'crayon', 'tea', 'toilet', 'walllamp', 'knife', 'floor', 'ceiling', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'couch', 'wall', 'oven', 'novel', 'coffe_maker', 'drinking_glass', 'tray', 'doorjamb', 'kettle', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'dishwasher', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and turn to the dishwasher, put the knife, glass, fork and plate into the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[TURNTO] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'walllamp', 'woman', 'drawing', 'phone', 'desk', 'man', 'toaster', 'laptop', 'mat', 'shower', 'electric_shaver', 'oil', 'form', 'pillow', 'dining_room', 'oven', 'shelf', 'bed', 'standingmirror', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'tooth_paste', 'scrabble']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'drawing', 'bathroom_counter', 'bedroom', 'keyboard', 'button', 'toaster', 'tvstand', 'food_noodles', 'pillow', 'check', 'television', 'oven_mitts', 'wine', 'curtain', 'rag', 'doorjamb', 'kettle', 'mouse', 'pencil', 'hairbrush', 'food_steak', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat; pet the cat by touching it", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'chair', 'bathroom', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'character', 'computer', 'mail', 'table_cloth', 'toaster', 'foundation', 'window', 'mousepad', 'bench', 'shelf', 'photoframe', 'bed', 'coffe_maker', 'cards', 'chessboard', 'controller', 'door', 'maindoor', 'bathtub', 'light', 'vase', 'laser_pointer']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk into living room, look at the phone when it ringing, grab the phone or pickup the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'plate', 'faucet', 'shower', 'couch', 'pillow', 'dining_room', 'wall', 'pasta', 'newspaper', 'novel', 'check', 'toilet_paper', 'food_food', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'controller', 'door', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I cross the room to the couch. I sit down and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'chair', 'freezer', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'button', 'mat', 'shoe-shine_kit', 'pillow', 'wall', 'oven', 'tape', 'bed', 'photoframe', 'novel', 'food_food', 'curtain', 'coffee_pot', 'mouse', 'pencil', 'conditioner', 'sink', 'coffee']\nTask:Read to child\nresponse:\n{\n    \"thoughts\": Read a book out loud to a child,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'freezer', 'dough', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'toaster', 'mat', 'pillow', 'dining_room', 'wall', 'oven', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'curtain', 'centerpiece', 'doorjamb', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'bag', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'phone', 'button', 'mat', 'stereo', 'window', 'homework', 'foundation', 'couch', 'pillow', 'wall', 'toy', 'cup', 'light_bulb', 'television', 'razor', 'curtain', 'after_shave', 'doorjamb', 'rag', 'coffee_pot', 'laundry_detergent', 'bathtub', 'table']\nTask:Arrange furniture\nresponse:\n{\n    \"thoughts\": Move furniture where you want it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[PUSH] <coffee_table>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'tvstand', 'faucet', 'plate', 'mousepad', 'wall', 'powersocket', 'bench', 'oven', 'cutting_board', 'tape', 'cup', 'doorjamb', 'sink', 'light', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk over and pick up the phone", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'floor', 'character', 'keyboard', 'toaster', 'mat', 'faucet', 'stereo', 'couch', 'towel', 'wall', 'wall_clock', 'nightstand', 'cup', 'coffe_maker', 'kitchen_counter', 'shoes', 'cd', 'hairbrush', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I have had my lunch and walked towards the sink and applied soap to my hands and washed cleanly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'walllamp', 'ground_coffee', 'orchid', 'drawing', 'cpuscreen', 'floor', 'phone', 'ceiling', 'toaster', 'ceilinglamp', 'plate', 'tvstand', 'window', 'dish_soap', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'razor', 'tray', 'mouse', 'kitchen_counter', 'cupboard', 'bathtub', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'food_salt', 'cd_player', 'chair', 'freezer', 'bathroom', 'food_carrot', 'orchid', 'knifeblock', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'remote_control', 'computer', 'window', 'mousepad', 'wall', 'nightstand', 'light_bulb', 'check', 'food_onion', 'curtain', 'mouse', 'door', 'light', 'tooth_paste']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.,\n    \"actions\": ['[WALK] <cd_player>', '[SWITCHON] <cd_player>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'tea_bag', 'bookshelf', 'food_bacon', 'toaster', 'plate', 'tvstand', 'dish_soap', 'mousepad', 'couch', 'wall', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'food_food', 'toothbrush_holder', 'ice', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'clothes_dress', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'juice', 'microwave', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'envelope', 'mop_bucket', 'toaster', 'mat', 'faucet', 'dish_soap', 'couch', 'wall', 'cup', 'photoframe', 'toilet_paper', 'video_game_controller', 'wooden_spoon', 'tray', 'console', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'light']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[SWITCHOFF] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <cup>', '[GRAB] <cup>', '[SCRUB] <cup>', '[PUTOBJBACK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[SCRUB] <cup>', '[PUTOBJBACK] <cup>', '[PUTBACK] <sponge> <sink>', '[SWITCHON] <faucet>', '[RINSE] <cup>', '[GRAB] <cup>', '[FIND] <dishrack>', '[PUTBACK] <cup> <dishrack>', '[RINSE] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishrack>', '[RINSE] <sponge>', '[SWITCHOFF] <faucet>', '[GRAB] <sponge>', '[PUTOBJBACK] <sponge>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'toilet', 'freezer', 'walllamp', 'drawing', 'character', 'bedroom', 'toaster', 'deck_of_cards', 'faucet', 'shower', 'plate', 'window', 'mousepad', 'couch', 'candle', 'clothes_hat', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'tray', 'mouse', 'pencil', 'bathtub', 'bathroom_cabinet', 'table', 'dishwasher', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the bedroom. I walk to stereo. I turn on stereo.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'piano_bench', 'mat', 'shoe-shine_kit', 'balanceball', 'oil', 'clothes_hat', 'dining_room', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'mouthwash', 'doorjamb', 'towel_rack', 'vacuum_cleaner', 'door', 'home_office', 'light', 'soap']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on the light,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'microwave', 'toilet', 'cat', 'wallshelf', 'bathroom_counter', 'floor', 'dvd_player', 'bookshelf', 'stovefan', 'ceilinglamp', 'toaster', 'mat', 'shower', 'faucet', 'tvstand', 'couch', 'towel', 'wall', 'bench', 'cutting_board', 'bed', 'razor', 'television', 'curtain', 'kitchen_counter', 'dresser', 'home_office', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Pet Cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'tea', 'microwave', 'jelly', 'band-aids', 'freezer', 'bathroom', 'needle', 'ceiling', 'character', 'keyboard', 'mop_bucket', 'ceilinglamp', 'mat', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'oven', 'photoframe', 'coffe_maker', 'check', 'television', 'cards', 'laundry_detergent', 'clothes_dress', 'home_office', 'table']\nTask:Put umbrella away\nresponse:\n{\n    \"thoughts\": Place the umbrella in the umbrella bucket point first.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mop_bucket>']\n}\n", "input": "\nScene_Objects:['food_salt', 'walllamp', 'toothbrush', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'dvd_player', 'keyboard', 'toaster', 'plate', 'foundation', 'dish_soap', 'form', 'wall', 'bench', 'novel', 'coffe_maker', 'check', 'food_donut', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'instrument_guitar', 'dishwasher', 'soap', 'scrabble']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Wash a plate using the dishwasher in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'walllamp', 'drawing', 'orchid', 'food_egg', 'floor', 'ceiling', 'phone', 'character', 'remote_control', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'shower', 'couch', 'clothes_hat', 'wall', 'powersocket', 'basket_for_clothes', 'check', 'mouthwash', 'television', 'food_onion', 'doorjamb', 'controller', 'door', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control,\n    \"actions\": ['[WALK] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'freezer', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'keyboard', 'bookshelf', 'piano_bench', 'stovefan', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'toy', 'bed', 'doorjamb', 'diary', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the sofa, and lie down on the sofa.", "actions": ["[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['band-aids', 'trashcan', 'wallshelf', 'clothes_underwear', 'desk', 'keyboard', 'mail', 'computer', 'tvstand', 'shower', 'window', 'clothes_socks', 'couch', 'food_peanut_butter', 'pillow', 'basket_for_clothes', 'keys', 'coffe_maker', 'television', 'food_donut', 'tray', 'rag', 'headset', 'clothes_scarf', 'bathtub', 'sink', 'light', 'dishwasher', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'detergent', 'knifeblock', 'phone', 'bookshelf', 'computer', 'mop_bucket', 'brush', 'clothes_socks', 'mousepad', 'dining_room', 'nightstand', 'basket_for_clothes', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'bathtub', 'sink', 'home_office', 'napkin', 'dishwasher', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "\"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\"", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[TURNTO] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'microwave', 'chair', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'computer', 'table_cloth', 'tvstand', 'shower', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'television', 'tray', 'doorjamb', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'board_game']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk into living room, walk to light, turn on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['walllamp', 'detergent', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'shower', 'stereo', 'couch', 'pillow', 'wall', 'oven', 'mouthwash', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'instrument_guitar', 'table', 'light', 'dishwasher', 'colander']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "lie in the sofa, relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'microwave', 'freezer', 'walllamp', 'bathroom', 'toothbrush', 'orchid', 'hanger', 'longboard', 'phone', 'desk', 'computer', 'food_bacon', 'table_cloth', 'clothes_jacket', 'brush', 'ceilinglamp', 'faucet', 'foundation', 'mousepad', 'clothes_pants', 'pasta', 'light_bulb', 'toothbrush_holder', 'coin', 'coffee_filter', 'towel_rack', 'hair', 'vase', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'beer', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'food_cake', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'photoframe', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'door', 'bathtub', 'bathroom_cabinet', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Sit on sofa. Lie down on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'food_ice_cream', 'microwave', 'freezer', 'ground_coffee', 'nail_polish', 'milk', 'hanger', 'floor', 'ceiling', 'toaster', 'ceilinglamp', 'faucet', 'clothes_hat', 'wall', 'wall_clock', 'oven', 'cutting_board', 'shelf', 'coffe_maker', 'check', 'television', 'coffee_filter', 'water', 'curtain', 'kitchen_counter', 'maindoor', 'sink', 'vase', 'soap']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find coffee maker. Find coffee filter. Put it in the coffee maker. Find ground coffee. Put it in the coffee maker. Find Water. Put it in the coffe maker. Switch on the coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[FIND] <water>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['water_glass', 'tablelamp', 'crayon', 'freezer', 'knifeblock', 'bathroom_counter', 'shampoo', 'ceiling', 'thread', 'faucet', 'shower', 'mousepad', 'food_kiwi', 'nightstand', 'wall', 'bench', 'powersocket', 'cup', 'clothes_gloves', 'novel', 'food_food', 'coffe_maker', 'toothbrush_holder', 'water', 'doorjamb', 'mouse', 'shoes', 'dresser', 'cupboard', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drinking a glass of water", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'freezer', 'walllamp', 'ground_coffee', 'bathroom', 'ceilingfan', 'clothes_underwear', 'bathroom_counter', 'character', 'thread', 'piano_bench', 'toaster', 'deck_of_cards', 'homework', 'couch', 'dustpan', 'powersocket', 'newspaper', 'clothes_gloves', 'novel', 'coffe_maker', 'doorjamb', 'chessboard', 'cd', 'napkin', 'home_office', 'instrument_guitar', 'clothes_skirt', 'light']\nTask:Close door\nresponse:\n{\n    \"thoughts\": Push the door shut until it latches.,\n    \"actions\": ['[RUN] <home_office>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'bathroom', 'detergent', 'cpuscreen', 'ceiling', 'computer', 'toaster', 'brush', 'laptop', 'faucet', 'mousepad', 'clothes_pants', 'dirt', 'powersocket', 'oven', 'cup', 'bed', 'food_food', 'television', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'chessboard', 'laundry_detergent', 'sink', 'dishwasher', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[PLUGIN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'bookmark', 'toaster', 'mat', 'tvstand', 'faucet', 'shower', 'laptop', 'mousepad', 'wall', 'oven', 'powersocket', 'towel_rack', 'kitchen_counter', 'headset', 'door', 'sink', 'bathtub', 'home_office', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I go to the bathroom. I locate the toilet. I walk to the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[WALK] <toilet>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'wallshelf', 'clothes_underwear', 'floor', 'desk', 'mail', 'button', 'stovefan', 'toaster', 'mousepad', 'soap', 'pillow', 'dining_room', 'photoframe', 'television', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'controller', 'bathtub', 'bathroom_cabinet', 'colander', 'clothes_shirt', 'coffee']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find a cat in the kitchen and touch it", "actions": ["[WALK] <dining_room>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'measuring_cup', 'microwave', 'chair', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'cleaning_solution', 'keyboard', 'mop_bucket', 'stovefan', 'faucet', 'couch', 'clothes_pants', 'coffe_maker', 'mirror', 'curtain', 'rag', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'dishwasher']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['sheets', 'jelly', 'freezer', 'ground_coffee', 'walllamp', 'trashcan', 'floor', 'ceiling', 'bookshelf', 'computer', 'toaster', 'stovefan', 'mat', 'mousepad', 'couch', 'clothes_hat', 'dining_room', 'wall', 'nightstand', 'cup', 'light_bulb', 'coffe_maker', 'coffee_filter', 'water', 'coffee_pot', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making Coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <water>", "[GRAB] <water>", "[OPEN] <coffe_maker>", "[POUR] <water> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <coffee_pot>", "[GRAB] <coffee_pot>", "[PUTBACK] <coffee_pot> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'walllamp', 'milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'candle', 'pillow', 'wall', 'cup', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'bathroom_cabinet', 'teeth']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sheets', 'ground_coffee', 'walllamp', 'trashcan', 'drawing', 'orchid', 'needle', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'toaster', 'ceilinglamp', 'couch', 'pillow', 'wall', 'nightstand', 'check', 'food_food', 'television', 'food_onion', 'doorjamb', 'pencil', 'door', 'home_office', 'table', 'light', 'coffee']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Walk to the sofa and sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'ground_coffee', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'shampoo', 'character', 'computer', 'ceilinglamp', 'mat', 'iron', 'plate', 'window', 'couch', 'dining_room', 'wall', 'powersocket', 'basket_for_clothes', 'photoframe', 'newspaper', 'video_game_controller', 'television', 'centerpiece', 'doorjamb', 'towel_rack', 'door', 'bathroom_cabinet', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on a couch, grab a book, open it, and scan the pages,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'dough', 'knifeblock', 'wallshelf', 'floor', 'phone', 'shampoo', 'desk', 'keyboard', 'bookmark', 'mat', 'shoe-shine_kit', 'floor_lamp', 'mousepad', 'pillow', 'powersocket', 'novel', 'curtain', 'centerpiece', 'kitchen_counter', 'door', 'bathtub', 'home_office', 'table', 'clothes_skirt', 'colander']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Turn a light on and read a book in the bedroom.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'detergent', 'knifeblock', 'phone', 'bookshelf', 'computer', 'mop_bucket', 'brush', 'clothes_socks', 'mousepad', 'dining_room', 'nightstand', 'basket_for_clothes', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'bathtub', 'sink', 'home_office', 'napkin', 'dishwasher', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['walllamp', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'alcohol', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'bench', 'photoframe', 'food_food', 'television', 'curtain', 'doorjamb', 'door', 'sink', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room, find the television, switch on the television, find the sofa, walk to sofa, sit in sofa, watch the television", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'dog', 'freezer', 'orchid', 'drawing', 'wallshelf', 'ceiling', 'character', 'mail', 'clothes_jacket', 'computer', 'mat', 'faucet', 'window', 'closetdrawer', 'nightstand', 'wall_clock', 'oven', 'basket_for_clothes', 'coffe_maker', 'mouse', 'dresser', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'light', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Pick up my dirty clothes, load the washer, add detergent, turn on washer,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[RUN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sponge', 'toilet', 'orchid', 'drawing', 'ceiling', 'bedroom', 'computer', 'mat', 'shoe-shine_kit', 'stereo', 'ceilinglamp', 'window', 'tvstand', 'food_orange', 'nightstand', 'bench', 'bed', 'coffe_maker', 'food_food', 'television', 'doorjamb', 'mouse', 'pencil', 'dresser', 'pajamas', 'cupboard', 'bathtub', 'table', 'clothes_shirt', 'coffee']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'longboard', 'character', 'bedroom', 'computer', 'faucet', 'homework', 'clothes_socks', 'food_noodles', 'dining_room', 'photoframe', 'check', 'television', 'after_shave', 'rag', 'lighter', 'towel_rack', 'kitchen_counter', 'dresser', 'clothes_dress', 'home_office', 'light', 'laser_pointer', 'scrabble']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office. I find the telephone. I pick up the telephone handset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'chair', 'walllamp', 'knifeblock', 'drawing', 'floor', 'phone', 'keyboard', 'toaster', 'stovefan', 'plate', 'shower', 'tvstand', 'dish_soap', 'couch', 'wall', 'coffe_maker', 'video_game_controller', 'food_onion', 'rag', 'towel_rack', 'mouse', 'kettle', 'kitchen_counter', 'cupboard', 'door', 'sink', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'measuring_cup', 'jelly', 'chair', 'freezer', 'hanger', 'ceiling', 'thread', 'computer', 'mail', 'bookmark', 'deck_of_cards', 'faucet', 'mousepad', 'dining_room', 'broom', 'wall', 'wall_clock', 'box', 'razor', 'doorjamb', 'mouse', 'pencil', 'hairbrush', 'door', 'napkin', 'table', 'light', 'dirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter the Home office, then walk to the phone, pickup the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'mop', 'walllamp', 'detergent', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'toaster', 'stereo', 'window', 'wall', 'bed', 'novel', 'coffe_maker', 'coffee_filter', 'water', 'kitchen_counter', 'shoes', 'dresser', 'cupboard', 'sink', 'bathtub', 'spoon', 'light', 'dishwasher', 'table', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Go to the kitchen and make coffee.", "actions": ["[RUN] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[PUTOBJBACK] <coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'nail_polish', 'bathroom', 'cat', 'dough', 'toothbrush', 'desk', 'button', 'stovefan', 'shower', 'faucet', 'window', 'pillow', 'nightstand', 'wall_clock', 'oven', 'cutting_board', 'novel', 'check', 'television', 'toothbrush_holder', 'lighter', 'kettle', 'chessboard', 'dresser', 'sink', 'home_office', 'clothes_skirt']\nTask:Keep cats out of room\nresponse:\n{\n    \"thoughts\": Stand in the doorway and create a loud noise so the cats will stay out.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>']\n}\n", "input": "\nScene_Objects:['spectacles', 'juice', 'food_salt', 'chair', 'freezer', 'ground_coffee', 'blow_dryer', 'nail_polish', 'food_egg', 'wallshelf', 'floor', 'ceiling', 'character', 'keyboard', 'computer', 'mop_bucket', 'toaster', 'ceilinglamp', 'clothes_hat', 'wall', 'food_food', 'food_donut', 'tray', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'laser_pointer']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Grab groceries and walk to the fridge. Open fridge and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'centerpiece', 'mouse', 'cupboard', 'sink', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Go find a computer, turn it on, click google, and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'food_bacon', 'mat', 'plate', 'faucet', 'shower', 'wall', 'nightstand', 'oven', 'toy', 'wooden_spoon', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet', 'bathtub']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walk into the kitchen. Walk up to the sink. Turn on the faucet. Pick up the plate. Wash the plate. Rinse the plate.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>", "[RINSE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'bathroom', 'orchid', 'food_egg', 'cpuscreen', 'floor', 'phone', 'bedroom', 'desk', 'stovefan', 'toaster', 'ceilinglamp', 'shower', 'mat', 'food_peanut_butter', 'wall', 'pasta', 'powersocket', 'oven', 'cup', 'bed', 'after_shave', 'food_onion', 'chessboard', 'dresser', 'cupboard', 'clothes_scarf', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my living room, find my cat, scrub it using my hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'stovefan', 'faucet', 'food_jam', 'wall', 'nightstand', 'oven', 'cup', 'photoframe', 'light_bulb', 'bed', 'clothes_gloves', 'television', 'water', 'tray', 'lighter', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'colander']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, grab a cup, fill with water and then drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[FIND] <sink>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'bathroom', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'toaster', 'stovefan', 'ceilinglamp', 'picture', 'tooth_paste', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'check', 'television', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'home_office', 'bathroom_cabinet', 'board_game']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'microwave', 'toilet', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'food_cake', 'food_bacon', 'table_cloth', 'bookmark', 'ceilinglamp', 'tvstand', 'faucet', 'couch', 'pillow', 'bowl', 'dining_room', 'wall', 'coffe_maker', 'doorjamb', 'door', 'home_office', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room,find the sofa,sat down in the sofa,take some relaxation.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'purse', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'shower', 'wall', 'powersocket', 'bench', 'cup', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'door', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Grasp light switch, flip light switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[TURNTO] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'measuring_cup', 'chair', 'walllamp', 'bathroom', 'drawing', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'food_cake', 'toaster', 'shower', 'picture', 'electric_shaver', 'window', 'nightstand', 'basket_for_clothes', 'bed', 'food_food', 'curtain', 'doorjamb', 'cupboard', 'napkin', 'instrument_guitar', 'bathroom_cabinet', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room, sit down by the cat, put out hand, stroke the cat from head to tail", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'scissors', 'measuring_cup', 'toilet', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'remote_control', 'purse', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'form', 'dining_room', 'wall', 'powersocket', 'television', 'curtain', 'food_onion', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to living room. Find remote control. Look at remote control. Push power on button. Find sofa. Sit down on sofa. Watch Television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'band-aids', 'walllamp', 'blow_dryer', 'detergent', 'bathroom', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'stereo', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'clothes_gloves', 'water', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'napkin']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'drawing', 'cpuscreen', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'alcohol', 'laptop', 'plate', 'faucet', 'homework', 'mousepad', 'dining_room', 'check', 'food_food', 'television', 'tray', 'wooden_spoon', 'mouse', 'shoes', 'conditioner', 'maindoor', 'napkin', 'home_office', 'soap', 'instrument_violin']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I turn on computer. I use keyboard to type. I open browser. I search.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'coffee_table', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'centerpiece', 'door', 'bathroom_cabinet', 'table', 'light', 'board_game']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <coffee_table>", "[PUTBACK] <remote_control> <coffee_table>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'knife', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'character', 'ceilinglamp', 'mat', 'shower', 'food_jam', 'dining_room', 'wall', 'bench', 'cutting_board', 'food_vegetable', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'table', 'light']\nTask:Chop vegetables\nresponse:\n{\n    \"thoughts\": Cut up veggies with knife,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[FIND] <cutting_board>', '[GRAB] <cutting_board>', '[CLOSE] <kitchen_cabinet>', '[WALK] <table>', '[PUTBACK] <knife> <table>', '[PUTBACK] <cutting_board> <table>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <food_vegetable> <cutting_board>', '[GRAB] <knife>', '[CUT] <food_vegetable>', '[PUTOBJBACK] <knife>']\n}\n", "input": "\nScene_Objects:['scissors', 'tablelamp', 'chair', 'walllamp', 'trashcan', 'drawing', 'knifeblock', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'computer', 'faucet', 'foundation', 'couch', 'wall', 'cup', 'coffe_maker', 'razor', 'tray', 'wooden_spoon', 'towel_rack', 'cupboard', 'laundry_detergent', 'sink', 'bathtub', 'bathroom_cabinet', 'dishwasher', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Going to the sink and scrubbing hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'ground_coffee', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'microwave', 'toilet', 'band-aids', 'walllamp', 'trashcan', 'knifeblock', 'drawing', 'ceiling', 'desk', 'stovefan', 'faucet', 'mousepad', 'wall', 'nightstand', 'oven', 'check', 'food_food', 'water', 'tray', 'doorjamb', 'kitchen_counter', 'cd', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'soap', 'laser_pointer']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'hanger', 'bathroom_counter', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'plate', 'tvstand', 'faucet', 'shower', 'dish_soap', 'mousepad', 'wall', 'wall_clock', 'oven', 'photoframe', 'bed', 'television', 'tray', 'towel_rack', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": walk into kitchen, find dishwasher, put plates in dishwasher, put soap in dishwasher, turn on dishwasher,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'foundation', 'couch', 'food_orange', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'console', 'door', 'home_office', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go into the living room. Walk over to the sofa. Lie down on the sofa. Turn on and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sheets', 'drawing', 'cpuscreen', 'ceiling', 'desk', 'bookmark', 'mat', 'plate', 'stereo', 'deck_of_cards', 'mousepad', 'couch', 'clothes_hat', 'clothes_pants', 'bench', 'clothes_gloves', 'video_game_controller', 'curtain', 'towel_rack', 'cd', 'dishrack', 'chessboard', 'pajamas', 'laundry_detergent', 'maindoor', 'door', 'sink', 'bathroom_cabinet', 'clothes_shirt']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the radio; switch on the radio and listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'cat', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'character', 'bedroom', 'alcohol', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'couch', 'food_orange', 'wall', 'photoframe', 'coffe_maker', 'check', 'curtain', 'doorjamb', 'shaving_cream', 'kitchen_counter', 'cupboard', 'sink', 'coffee_cup', 'dishwasher', 'table', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to the kitchen at the coffee maker, grab coffee, pour coffee into coffee maker, put coffee into coffee maker, push coffee maker's button, wait to complete, grab coffee cup", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[OPEN] <coffe_maker>", "[POUR] <coffee> <coffe_maker>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'cat', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'character', 'keyboard', 'remote_control', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'oil', 'couch', 'pillow', 'wall', 'check', 'television', 'curtain', 'door', 'home_office', 'table', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Pickup remote control for the living room TV and push the change channel button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'ground_coffee', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'table_cloth', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'couch', 'candle', 'dustpan', 'pillow', 'dining_room', 'wall', 'powersocket', 'cup', 'photoframe', 'tray', 'curtain', 'door', 'bathtub', 'sink', 'light', 'table', 'dishwasher', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into my room and turn the radio on so that I can listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'laundry_detergent', 'door', 'table', 'light', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn tv on, watch TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'keyboard', 'computer', 'clothes_jacket', 'ceilinglamp', 'faucet', 'shower', 'window', 'towel', 'wall', 'newspaper', 'curtain', 'towel_rack', 'pencil', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[PUTOBJBACK] <soap>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'toaster', 'mat', 'plate', 'shower', 'iron', 'dish_soap', 'closetdrawer', 'pillow', 'wall', 'oven', 'basket_for_clothes', 'coffe_maker', 'check', 'tray', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'dishwasher', 'board_game', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>', '[GRAB] <plate>', '[WIPE] <plate>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[PUTOBJBACK] <dishrack>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'laundry_detergent', 'door', 'table', 'light', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "turn tv on, watch TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'orchid', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'desk', 'food_cake', 'stovefan', 'ceilinglamp', 'toaster', 'food_orange', 'wall', 'pasta', 'oven', 'tape', 'tray', 'wooden_spoon', 'centerpiece', 'kitchen_counter', 'cupboard', 'napkin', 'dishwasher', 'clothes_shirt']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Walk into the kitchen. Open the fridge. Take out the chicken. Open the kitchen cabinet. Take out the frying pan. Put the chicken in the frying pan. Put the frying pan on the stove. Turn on the stove. Cook the chicken.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[CLOSE] <freezer>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[CLOSE] <kitchen_cabinet>', '[WALK] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'drawing', 'knife', 'wallshelf', 'orchid', 'floor', 'bookshelf', 'mail', 'mat', 'tvstand', 'shower', 'window', 'mousepad', 'candle', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'food_food', 'coffe_maker', 'food_rice', 'tray', 'shoes', 'door', 'clothes_scarf', 'bathtub', 'bathroom_cabinet', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I would go lie down in my bed and open the book and start reading.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'trashcan', 'desk', 'keyboard', 'stovefan', 'ceilinglamp', 'mat', 'toaster', 'stereo', 'window', 'couch', 'dining_room', 'nightstand', 'wall', 'cutting_board', 'bed', 'photoframe', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'hairdryer', 'bathtub', 'bathroom_cabinet', 'home_office', 'dishwasher', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk into room, grab phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'microwave', 'chair', 'knifeblock', 'drawing', 'keyboard', 'button', 'stovefan', 'toaster', 'shower', 'window', 'dustpan', 'pillow', 'dining_room', 'nightstand', 'tape', 'novel', 'television', 'tray', 'curtain', 'after_shave', 'doorjamb', 'kitchen_counter', 'cupboard', 'clothes_dress', 'sink', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I pick up my phone. I touch it to navigate to the web browser. I can swipe and tap the screen to visit different websites. Then I read or watch the content within those websites.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[TOUCH] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'nail_polish', 'food_carrot', 'drawing', 'knife', 'floor', 'ceiling', 'desk', 'mat', 'stereo', 'shower', 'clothes_socks', 'candle', 'pillow', 'dining_room', 'nightstand', 'bench', 'newspaper', 'check', 'video_game_controller', 'doorjamb', 'cards', 'chessboard', 'pajamas', 'clothes_dress', 'table', 'light', 'board_game', 'laser_pointer', 'coffee']\nTask:Play games\nresponse:\n{\n    \"thoughts\": I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <board_game>', '[FIND] <board_game>', '[GRAB] <board_game>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <board_game> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'bathroom', 'cat', 'milk', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'bookshelf', 'computer', 'plate', 'stereo', 'clothes_pants', 'pillow', 'oven', 'food_food', 'razor', 'towel_rack', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'shoes', 'dresser', 'cupboard', 'bathtub', 'light', 'dishwasher', 'laser_pointer']\nTask:Give milk to cat\nresponse:\n", "output": {"thoughts": "Grab milk and pour it into the plate so that the cat can drink", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[POINTAT] <cat>", "[PULL] <cat>", "[FIND] <milk>", "[GRAB] <milk>", "[FIND] <plate>", "[POUR] <milk> <plate>", "[TOUCH] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'chair', 'freezer', 'drawing', 'knife', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'toaster', 'tvstand', 'shoe-shine_kit', 'faucet', 'window', 'mousepad', 'couch', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'curtain', 'coffee_pot', 'towel_rack', 'sink', 'light', 'laser_pointer']\nTask:Gaze out window\nresponse:\n{\n    \"thoughts\": Look at the yard from the window,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <window>', '[FIND] <window>', '[TURNTO] <window>', '[LOOKAT] <window>']\n}\n", "input": "\nScene_Objects:['glue', 'food_ice_cream', 'tea', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'drawing', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'dining_room', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'cupboard', 'light', 'dishwasher']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the television. Switch on the television. Find the remote control. Grab the remote control. Point at television. Find the button. Push button.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'tea', 'band-aids', 'walllamp', 'nail_polish', 'knifeblock', 'bills', 'cpuscreen', 'hanger', 'shampoo', 'phone', 'floor', 'bedroom', 'desk', 'toaster', 'stereo', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bowl', 'dining_room', 'bench', 'newspaper', 'clothes_gloves', 'coffe_maker', 'curtain', 'diary', 'hairbrush', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'nail_polish', 'food_carrot', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'stovefan', 'mat', 'tvstand', 'homework', 'window', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'oven', 'bench', 'bed', 'novel', 'coffe_maker', 'toothbrush_holder', 'towel_rack', 'home_office', 'instrument_violin']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the bedroom. Grab the book out of the nightstand. Lay down in bed. Open the book. Read the book.", "actions": ["[WALK] <bedroom>", "[WALK] <nightstand>", "[FIND] <nightstand>", "[OPEN] <nightstand>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[TURNTO] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'tablelamp', 'crayon', 'tea', 'mop', 'ground_coffee', 'detergent', 'drawing', 'floor', 'phone', 'ceilinglamp', 'laptop', 'shower', 'iron', 'window', 'oil', 'couch', 'pillow', 'wall', 'oven', 'cup', 'coffe_maker', 'toilet_paper', 'television', 'door', 'sink', 'table', 'scrabble', 'dirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Bring book, lie on the sofa, read book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[LIE] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'food_salt', 'walllamp', 'bathroom', 'trashcan', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'tvstand', 'shower', 'faucet', 'towel', 'wall', 'cup', 'keys', 'food_food', 'television', 'water', 'curtain', 'lighter', 'kitchen_counter', 'shoes', 'door', 'table', 'light', 'soap', 'teeth']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "go to a bathroom, open the shower, let water wash your body", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'dog', 'chair', 'bathroom', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'mat', 'ceilinglamp', 'shower', 'candle', 'wall', 'bed', 'clothes_gloves', 'food_food', 'water', 'food_rice', 'wooden_spoon', 'curtain', 'console', 'dishrack', 'pajamas', 'light', 'clothes_skirt', 'soap', 'dirt']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'toilet', 'chair', 'knife', 'knifeblock', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'bookmark', 'plate', 'towel', 'pillow', 'photoframe', 'novel', 'coffe_maker', 'check', 'coffee_filter', 'curtain', 'doorjamb', 'towel_rack', 'sink', 'home_office', 'table', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "BEFORE EATING", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <towel>", "[WIPE] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'window', 'toaster', 'clothes_pants', 'broom', 'powersocket', 'bench', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'dresser', 'hair', 'door', 'sink', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Get some clothes together that make an appropriate outfit.  Put them on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'freezer', 'walllamp', 'toothbrush', 'drawing', 'knifeblock', 'bathroom_counter', 'shampoo', 'floor', 'bookshelf', 'food_cake', 'mail', 'stovefan', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'wall', 'photoframe', 'toothbrush_holder', 'towel_rack', 'pencil', 'door', 'sink', 'bathroom_cabinet', 'table', 'home_office', 'tooth_paste', 'coffee']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "getting ready for bed", "actions": ["[WALK] <bathroom_counter>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <toothbrush> <bathroom_counter>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[GRAB] <toothbrush>", "[SCRUB] <toothbrush>", "[RINSE] <toothbrush>", "[FIND] <toothbrush_holder>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ground_coffee', 'nail_polish', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'button', 'mail', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'tape', 'television', 'curtain', 'food_onion', 'cupboard', 'table', 'shoe_rack']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to the living room, sit on the couch, find the remote, switch on the TV and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'toilet', 'chair', 'bathroom', 'orchid', 'knifeblock', 'floor', 'phone', 'desk', 'keyboard', 'food_cake', 'foundation', 'floor_lamp', 'pillow', 'nightstand', 'powersocket', 'box', 'basket_for_clothes', 'razor', 'tray', 'doorjamb', 'cupboard', 'bathtub', 'spoon', 'home_office', 'table', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room, find lamp, switch on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'walllamp', 'toothbrush', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'toaster', 'ceilinglamp', 'window', 'homework', 'food_noodles', 'couch', 'candle', 'pillow', 'dining_room', 'wall', 'powersocket', 'toy', 'food_food', 'toilet_paper', 'tray', 'curtain', 'food_onion', 'shoes', 'door', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['fork', 'measuring_cup', 'walllamp', 'knifeblock', 'knife', 'drawing', 'floor', 'ceiling', 'character', 'toaster', 'mat', 'plate', 'shower', 'dish_soap', 'clothes_socks', 'wall', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dishrack', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet', 'instrument_guitar', 'dishwasher', 'bathtub']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <dishrack>", "[PUTBACK] <fork> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishrack>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishrack>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'window', 'toaster', 'clothes_pants', 'broom', 'powersocket', 'bench', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'dresser', 'hair', 'door', 'sink', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Get some clothes together that make an appropriate outfit.  Put them on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'remote_control', 'computer', 'button', 'toaster', 'mat', 'shower', 'food_peanut_butter', 'food_kiwi', 'nightstand', 'oven', 'photoframe', 'television', 'oven_mitts', 'tray', 'kettle', 'cards', 'pencil', 'kitchen_counter', 'towel_rack', 'bathtub', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk into living room. Find television remote control. Pick up remote control. Press button.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'food_carrot', 'woman', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'child', 'bed', 'television', 'curtain', 'towel_rack', 'dresser', 'table']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Sitting and watching  a movie with family,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TURNTO] <child>', '[LOOKAT] <child>', '[GREET] <child>', '[FIND] <woman>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>', '[WALK] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'juice', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'drawing', 'cpuscreen', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'bookshelf', 'laptop', 'tvstand', 'faucet', 'electric_shaver', 'dining_room', 'wall', 'nightstand', 'oven', 'photoframe', 'novel', 'water', 'kitchen_counter', 'cupboard', 'hairdryer', 'sink', 'door']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Opening the kitchen cabinet to grab water glass and pour water from the faucet into it to drink.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[CLOSE] <kitchen_cabinet>", "[FIND] <sink>", "[PUTBACK] <water_glass> <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[SWITCHOFF] <faucet>", "[GRAB] <water_glass>", "[FIND] <water>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'stamp', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Pay bills\nresponse:\n{\n    \"thoughts\": Check my bills, on the computer, open the appropriate sites, pay my due bills,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'newspaper', 'television', 'curtain', 'doorjamb', 'door', 'home_office', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room. look at remote control. grab remote control. plug in television. switch on television. lie in sofa. point at remote control. watch television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[LIE] <couch>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Draft home\nresponse:\n{\n    \"thoughts\": Using mathematics and CAD software design a home,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'freezer', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'pasta', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I come home and go to the sofa for turn off the tv and relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'chair', 'freezer', 'bathroom', 'food_butter', 'drawing', 'food_egg', 'floor', 'computer', 'mat', 'tvstand', 'stereo', 'laptop', 'clothes_pants', 'towel', 'wall', 'nightstand', 'tape', 'bed', 'food_food', 'coffe_maker', 'television', 'oven_mitts', 'tray', 'curtain', 'diary', 'cards', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Turn on radio, put on favorite channel, listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['mop', 'microwave', 'cd_player', 'walllamp', 'drawing', 'orchid', 'bills', 'floor', 'ceiling', 'desk', 'tea_bag', 'button', 'piano_bench', 'stovefan', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'ice', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'door', 'bathtub', 'home_office', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room, walk to cd player, turn on cd player, press play button, walk to sofa, sit in sofa and listen to music", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <button>", "[PUSH] <button>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'freezer', 'walllamp', 'knifeblock', 'wallshelf', 'clothes_underwear', 'ceiling', 'bookshelf', 'keyboard', 'ceilinglamp', 'plate', 'stereo', 'clothes_socks', 'clothes_pants', 'pillow', 'bowl', 'powersocket', 'cutting_board', 'basket_for_clothes', 'razor', 'coin', 'wine', 'curtain', 'pencil', 'kitchen_counter', 'cd', 'dresser', 'laundry_detergent', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'tvstand', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'cup', 'television', 'doorjamb', 'mouse', 'shoes', 'door', 'sink', 'napkin', 'clothes_scarf', 'table', 'light', 'dishwasher', 'colander']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room ,find sofa ,sit on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'walllamp', 'dough', 'woman', 'knifeblock', 'bathroom_counter', 'floor', 'character', 'desk', 'tea_bag', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'wall', 'child', 'coffe_maker', 'video_game_controller', 'television', 'towel_rack', 'kettle', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'instrument_violin', 'clothes_shirt']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Go to the kitchen, greet your family members,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <woman>', '[GREET] <woman>', '[FIND] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['beer', 'bathroom', 'orchid', 'knifeblock', 'food_egg', 'clothes_underwear', 'shampoo', 'bedroom', 'tvstand', 'shoe-shine_kit', 'pillow', 'dining_room', 'bench', 'powersocket', 'coffe_maker', 'video_game_controller', 'television', 'food_rice', 'tray', 'after_shave', 'food_onion', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'comb', 'cupboard', 'bathroom_cabinet', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TOUCH] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'tea', 'walllamp', 'cat', 'trashcan', 'food_butter', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'shower', 'closetdrawer', 'couch', 'dining_room', 'oven', 'bench', 'bed', 'food_food', 'television', 'food_onion', 'doorjamb', 'cupboard', 'sink', 'bathtub', 'soap', 'laser_pointer']\nTask:Walk to room\nresponse:\n{\n    \"thoughts\": Leave the living room and enter the hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'trashcan', 'drawing', 'floor', 'bedroom', 'keyboard', 'stovefan', 'tvstand', 'homework', 'window', 'microphone', 'food_orange', 'pillow', 'dining_room', 'powersocket', 'bench', 'newspaper', 'bed', 'coffe_maker', 'food_food', 'tray', 'pencil', 'kitchen_counter', 'spoon', 'home_office', 'bathroom_cabinet']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter the Home office, then walk to the phone, pickup the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'walllamp', 'needle', 'wallshelf', 'cpuscreen', 'hanger', 'bedroom', 'keyboard', 'bookmark', 'stovefan', 'shower', 'oil', 'dining_room', 'wall', 'powersocket', 'bench', 'love_seat', 'standingmirror', 'novel', 'coffe_maker', 'check', 'after_shave', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'maindoor', 'bathtub', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'clothes_underwear', 'floor', 'ceiling', 'thread', 'computer', 'bookmark', 'toaster', 'stovefan', 'couch', 'wall', 'powersocket', 'cup', 'food_food', 'razor', 'water', 'tray', 'centerpiece', 'pencil', 'kitchen_counter', 'chessboard', 'cupboard', 'clothes_scarf', 'light', 'table', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <freezer>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <cup>", "[POUR] <water> <cup>", "[FIND] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'microwave', 'chair', 'freezer', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'stereo', 'picture', 'mat', 'closetdrawer', 'couch', 'dining_room', 'wall_clock', 'bench', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'home_office', 'instrument_guitar']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the living room and then I walk to the light switch in there and turn the lights on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'folder', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'photoframe', 'food_food', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'towel_rack', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the television. Walk to the television. Switch on the television. Find the sofa. Walk to the sofa. Sit on the sofa. Find the remote control. Grab the remote control. Look at the television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[WALK] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['trashcan', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'tea_bag', 'keyboard', 'food_bacon', 'stovefan', 'tvstand', 'faucet', 'couch', 'microphone', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'oven_mitts', 'doorjamb', 'mouse', 'door', 'bathtub', 'sink', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'tablelamp', 'juice', 'cd_player', 'band-aids', 'walllamp', 'cpuscreen', 'bathroom_counter', 'ceiling', 'desk', 'stovefan', 'stereo', 'homework', 'food_orange', 'nightstand', 'bench', 'bed', 'newspaper', 'photoframe', 'coffe_maker', 'check', 'food_onion', 'curtain', 'doorjamb', 'dresser', 'door', 'sink', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk in the living room, then I go the the stereo and the cd player and turn them on.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'shoe_rack', 'food_salt', 'beer', 'measuring_cup', 'microwave', 'trashcan', 'food_carrot', 'orchid', 'drawing', 'bookshelf', 'envelope', 'alcohol', 'ceilinglamp', 'plate', 'food_jam', 'pillow', 'oven', 'newspaper', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'lighter', 'doorjamb', 'coffee_pot', 'towel_rack', 'cupboard', 'hairdryer', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find a cat, touch it gently,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'deck_of_cards', 'faucet', 'mousepad', 'pillow', 'wall', 'oven', 'box', 'cup', 'toilet_paper', 'rag', 'kitchen_counter', 'cupboard', 'sink', 'bag', 'bathroom_cabinet', 'bathtub', 'table', 'home_office']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk the kitchen and walk up to the kitchen cabinet. I open the cabinet and take out a cup. I walk to sink and I turn the faucet on using the cold water handle. I put my cup under the running water until it is full. I turn off the cold water handle and bring the cup to my mouth and take a drink.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <cup>", "[GRAB] <cup>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[WALK] <sink>", "[FIND] <faucet>", "[PUSH] <faucet>", "[PUTBACK] <cup> <sink>", "[GRAB] <cup>", "[SWITCHOFF] <faucet>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'bedroom', 'keyboard', 'tvstand', 'stereo', 'balanceball', 'picture', 'mousepad', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'photoframe', 'bed', 'toilet_paper', 'coffe_maker', 'cards', 'mouse', 'kitchen_counter', 'cd', 'vase']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'character', 'bedroom', 'desk', 'bookshelf', 'computer', 'mat', 'tvstand', 'window', 'pillow', 'wall', 'powersocket', 'tray', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'bathtub', 'light', 'teeth', 'shoe_rack']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "walk to bathroom , run to toilet", "actions": ["[WALK] <bathroom>", "[RUN] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'filing_cabinet', 'toilet', 'detergent', 'food_butter', 'orchid', 'needle', 'bills', 'stereo', 'window', 'dining_room', 'bench', 'oven', 'bed', 'novel', 'check', 'coffee_filter', 'tray', 'wooden_spoon', 'centerpiece', 'headset', 'conditioner', 'food_steak', 'napkin', 'home_office', 'soap', 'scrabble', 'dirt', 'coffee']\nTask:File documents\nresponse:\n{\n    \"thoughts\": I file all of my paperwork (including bills, receipts, documents, device instructions, etc.) in my black paperwork case that has many different files in order to organize.  I place each piece of paper in its' corresponding subject (as indicated by printed words on each tab).,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bills>', '[FIND] <bills>', '[GRAB] <bills>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTIN] <bills> <filing_cabinet>', '[FIND] <check>', '[GRAB] <check>', '[WALK] <filing_cabinet>', '[PUTIN] <check> <filing_cabinet>', '[FIND] <check>', '[GRAB] <check>', '[WALK] <filing_cabinet>', '[PUTIN] <check> <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['water_glass', 'tablelamp', 'juice', 'food_ice_cream', 'measuring_cup', 'chair', 'freezer', 'nail_polish', 'detergent', 'food_carrot', 'toothbrush', 'drawing', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'bookmark', 'toaster', 'couch', 'wall', 'keys', 'toilet_paper', 'television', 'kitchen_counter', 'headset', 'cupboard', 'door', 'home_office', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen ,find fridge ,open fridge ,find water  glass , drink from water glass", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'microwave', 'freezer', 'knife', 'food_egg', 'wallshelf', 'kitchen_cabinet', 'ceiling', 'character', 'desk', 'tea_bag', 'toaster', 'stovefan', 'plate', 'faucet', 'shower', 'mat', 'wall', 'oven', 'bench', 'coffe_maker', 'drinking_glass', 'tray', 'kitchen_counter', 'cupboard', 'door', 'sink', 'spoon', 'napkin', 'dishwasher']\nTask:Unload dishwasher\nresponse:\n{\n    \"thoughts\": Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[PUTBACK] <drinking_glass> <cupboard>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <fork> <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTIN] <knife> <kitchen_cabinet>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTIN] <spoon> <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'conditioner', 'door', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "For relaxing walk to living room. Find the sofa to relax. Arrange the sofa to relax. Sit on the sofa to relaxing.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <conditioner>", "[FIND] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'jelly', 'freezer', 'walllamp', 'bathroom', 'hanger', 'folder', 'bedroom', 'desk', 'tvstand', 'window', 'food_jam', 'microphone', 'dining_room', 'wall_clock', 'powersocket', 'bench', 'shelf', 'bed', 'video_game_controller', 'food_donut', 'water', 'doorjamb', 'cards', 'towel_rack', 'mouse', 'sink', 'colander', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the water, wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <water>', '[FIND] <water>', '[TURNTO] <water>', '[LOOKAT] <water>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'trashcan', 'orchid', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'bookshelf', 'clothes_jacket', 'faucet', 'couch', 'clothes_pants', 'form', 'wall', 'cutting_board', 'photoframe', 'check', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'clothes_dress', 'door', 'sink', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open cabinet and collect clothes and put them into washing machine and wash clothes", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'toilet', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'bowl', 'wall', 'television', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'light', 'table', 'board_game']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Turn on PC and log onto email account. Type email and send it. Switch off computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'shower', 'window', 'clothes_hat', 'dining_room', 'wall', 'bench', 'powersocket', 'photoframe', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'table', 'light', 'tooth_paste']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Pick up phone", "actions": ["[WALK] <table>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'mop', 'microwave', 'toilet', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'computer', 'piano_bench', 'ceilinglamp', 'plate', 'deck_of_cards', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'bed', 'novel', 'television', 'doorjamb', 'door', 'coffee_cup', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['food_salt', 'trashcan', 'toothbrush', 'drawing', 'knifeblock', 'wallshelf', 'bathroom_counter', 'ceiling', 'phone', 'bedroom', 'stovefan', 'tvstand', 'food_peanut_butter', 'nightstand', 'wall', 'oven', 'powersocket', 'coffe_maker', 'toothbrush_holder', 'after_shave', 'curtain', 'coffee_pot', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'bathtub', 'sink', 'home_office', 'dishwasher', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'stovefan', 'balanceball', 'mat', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'light', 'dishwasher']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, sit at desk, turn on computer, enter password, open application and begin work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'blender', 'microwave', 'toilet', 'detergent', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'character', 'keyboard', 'computer', 'brush', 'mat', 'laptop', 'faucet', 'ceilinglamp', 'window', 'dustpan', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'towel_rack', 'mouse', 'kitchen_counter', 'pajamas', 'dishwasher']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "go to bedroom, put pajamas on, lie in bed, sleep", "actions": ["[WALK] <bedroom>", "[WALK] <pajamas>", "[FIND] <pajamas>", "[GRAB] <pajamas>", "[PUTON] <pajamas>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'walllamp', 'detergent', 'orchid', 'drawing', 'cpuscreen', 'keyboard', 'envelope', 'computer', 'stovefan', 'faucet', 'electric_shaver', 'couch', 'candle', 'microphone', 'broom', 'nightstand', 'oven', 'bench', 'tape', 'cup', 'light_bulb', 'food_rice', 'wine', 'wooden_spoon', 'hair', 'maindoor', 'table', 'dishwasher', 'vase']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Get a broom or duster with long handle. Dust blades of ceiling fan.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <duster>', '[FIND] <duster>', '[GRAB] <duster>', '[PUTOBJBACK] <duster>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'toaster', 'mat', 'ceilinglamp', 'plate', 'stovefan', 'dish_soap', 'broom', 'wall', 'bench', 'food_food', 'coffe_maker', 'tray', 'wooden_spoon', 'towel_rack', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'table', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go the in kitchen then grab the plate on the table. I put the plate and dish soap in the dishwasher. Finally, I turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'freezer', 'blow_dryer', 'walllamp', 'nail_polish', 'bathroom', 'toothbrush', 'drawing', 'hanger', 'floor', 'shampoo', 'character', 'desk', 'bookshelf', 'purse', 'envelope', 'mat', 'ceilinglamp', 'window', 'towel', 'wall', 'powersocket', 'toy', 'cup', 'food_food', 'coffe_maker', 'sink', 'duster']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Apply duster to shelves and clean them until dust is gone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <duster>', '[FIND] <duster>', '[GRAB] <duster>', '[FIND] <filing_cabinet>', '[WIPE] <filing_cabinet>', '[PUTOBJBACK] <duster>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'knife', 'knifeblock', 'orchid', 'drawing', 'floor', 'phone', 'character', 'bedroom', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'faucet', 'shower', 'oil', 'dining_room', 'powersocket', 'newspaper', 'ice', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'door', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pick up the remote. I press button on remote.", "actions": ["[WALK] <bedroom>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'bookshelf', 'piano_bench', 'stovefan', 'ceilinglamp', 'faucet', 'electric_shaver', 'window', 'teeth', 'food_orange', 'pillow', 'wall', 'powersocket', 'food_food', 'tray', 'curtain', 'dresser', 'hair', 'controller', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk in to the bathroom. I find the soap. I find the faucet. I turn on the faucet. I wash my hands. I put the soap back. I rinse my hands. I turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'orchid', 'ceiling', 'desk', 'computer', 'food_bacon', 'stovefan', 'ceilinglamp', 'plate', 'shower', 'faucet', 'window', 'form', 'dining_room', 'oven', 'basket_for_clothes', 'keys', 'novel', 'food_food', 'coffe_maker', 'food_rice', 'tray', 'doorjamb', 'cards', 'mouse', 'pajamas', 'home_office', 'light', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the playback device you want to listen to. turn on. press play. enjoyu music.", "actions": ["[WALK] <bedroom>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'detergent', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'shoe-shine_kit', 'faucet', 'clothes_socks', 'closetdrawer', 'mousepad', 'dining_room', 'broom', 'wall', 'nightstand', 'photoframe', 'television', 'water', 'tray', 'cards', 'paper_towel', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. I turn on the water. I put soap on my hands. I rinse my hands. I turn off the water. I dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <water>', '[TURNTO] <water>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[TURNTO] <water>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>']\n}\n", "input": "\nScene_Objects:['walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'button', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'nightstand', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'pencil', 'dresser', 'cupboard', 'hair', 'light', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Go to the living room, Switch on the television. Find the remote and press the change channel button, change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[TOUCH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'filing_cabinet', 'scissors', 'chair', 'ceilingfan', 'dough', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'coffee_table', 'television', 'coin', 'wooden_spoon', 'curtain', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'folder', 'bookshelf', 'remote_control', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'door', 'bathroom_cabinet', 'light', 'table', 'tooth_paste']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'table_cloth', 'stovefan', 'mat', 'faucet', 'foundation', 'mousepad', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'toy', 'hands_both', 'tray', 'doorjamb', 'towel_rack', 'shaving_cream', 'door', 'sink', 'spoon', 'bathroom_cabinet', 'light', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I need to rinse my dentures off and put them back in my mouth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <teeth>', '[GRAB] <teeth>', '[WASH] <teeth>', '[PUTOBJBACK] <teeth>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <towel>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'table_cloth', 'tvstand', 'faucet', 'stereo', 'window', 'couch', 'form', 'wall', 'powersocket', 'light_bulb', 'razor', 'vacuum_cleaner', 'controller', 'door', 'home_office', 'light', 'dishwasher', 'bathroom_cabinet', 'colander']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bath room and go to the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'blow_dryer', 'bathroom', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'stamp', 'keyboard', 'computer', 'food_bacon', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'coffe_maker', 'wooden_spoon', 'towel_rack', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'controller', 'light', 'soap']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Plug in vacuum. Walk around with vacuum, covering all of the floor. Unplug vacuum.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['fork', 'spectacles', 'chair', 'freezer', 'walllamp', 'knife', 'wallshelf', 'floor', 'alcohol', 'bookmark', 'toaster', 'plate', 'candle', 'wall', 'nightstand', 'novel', 'coffe_maker', 'food_food', 'television', 'water', 'curtain', 'after_shave', 'diary', 'kitchen_counter', 'dresser', 'cupboard', 'bathtub', 'spoon', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "first i will take all the dishes to the dish washer and with the help of dishwasher and water i will clean all the dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water>", "[TURNTO] <water>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'sheets', 'food_salt', 'tea', 'blow_dryer', 'wallshelf', 'ceiling', 'character', 'food_cake', 'computer', 'tvstand', 'window', 'mousepad', 'cutting_board', 'basket_for_clothes', 'cup', 'newspaper', 'light_bulb', 'novel', 'photoframe', 'coin', 'food_donut', 'rag', 'cards', 'kitchen_counter', 'vacuum_cleaner', 'pajamas', 'maindoor', 'bag', 'creditcard']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Clean the welcome mat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <vacuum_cleaner>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['fork', 'crayon', 'microwave', 'chair', 'cloth_napkin', 'freezer', 'walllamp', 'drawing', 'knife', 'floor', 'phone', 'character', 'food_cake', 'mat', 'ceilinglamp', 'plate', 'window', 'mousepad', 'microphone', 'pillow', 'wall', 'nightstand', 'bench', 'cup', 'video_game_controller', 'doorjamb', 'controller', 'door', 'table', 'teeth']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk to dining room, walk to dining table, arrange chairs, set napkin, set plate, set knife, set fork and glass on table.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <chair>", "[PULL] <chair>", "[FIND] <cloth_napkin>", "[GRAB] <cloth_napkin>", "[PUTBACK] <cloth_napkin> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'filing_cabinet', 'tablelamp', 'mop', 'microwave', 'chair', 'dog', 'toothbrush', 'knifeblock', 'hanger', 'ceiling', 'bedroom', 'man', 'keyboard', 'ceilinglamp', 'stereo', 'shower', 'couch', 'food_jam', 'pillow', 'pasta', 'powersocket', 'coffe_maker', 'kettle', 'pencil', 'napkin', 'table', 'scrabble', 'creditcard']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": When someone knocks I n the door walk through the hallway and greet them.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'nail_polish', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'bedroom', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'mat', 'window', 'couch', 'dining_room', 'wall', 'tape', 'cup', 'bed', 'mouthwash', 'after_shave', 'dishrack', 'cupboard', 'door', 'sink', 'table', 'light', 'tooth_paste', 'clothes_shirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'desk', 'brush', 'mat', 'ceilinglamp', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'wall', 'powersocket', 'cup', 'photoframe', 'coffee_filter', 'food_rice', 'doorjamb', 'towel_rack', 'dresser', 'clothes_dress', 'door', 'light', 'table']\nTask:Change clothes\nresponse:\n{\n    \"thoughts\": Switch your clothes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTON] <clothes_dress>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'ceilinglamp', 'shoe-shine_kit', 'faucet', 'clothes_pants', 'dining_room', 'wall', 'tape', 'basket_for_clothes', 'bed', 'coffe_maker', 'chessboard', 'clothes_dress', 'door', 'home_office', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'toaster', 'closetdrawer', 'couch', 'food_jam', 'pillow', 'wall', 'cup', 'photoframe', 'coffee_table', 'novel', 'television', 'doorjamb', 'shoes', 'comb', 'dresser', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['blender', 'mop', 'walllamp', 'trashcan', 'wallshelf', 'floor', 'ceiling', 'folder', 'bedroom', 'tvstand', 'shower', 'foundation', 'window', 'mousepad', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'food_food', 'coffe_maker', 'television', 'food_donut', 'curtain', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to the bathroom and wash with the shower.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[WASH] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'toothbrush', 'drawing', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'mat', 'faucet', 'shower', 'food_noodles', 'form', 'wall', 'wall_clock', 'oven', 'tape', 'photoframe', 'keys', 'toothbrush_holder', 'rag', 'doorjamb', 'sink', 'bathroom_cabinet', 'instrument_guitar', 'dishwasher', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I walk in to the bathroom. I turn on the faucet. I find the toothbrush and toothpaste. I grab the toothbrush and toothpaste. I rinse off the toothbrush. I add toothpaste to the toothbrush. I put back the toothpaste. I wash my teeth with the toothbrush. I rinse off the toothbrush. I put the toothbrush back. I rinse my mouth out. I turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[RINSE] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>', '[PUTOBJBACK] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>', '[RINSE] <toothbrush>', '[PUTOBJBACK] <toothbrush>', '[RINSE] <teeth>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['toilet', 'detergent', 'knifeblock', 'knife', 'bills', 'clothes_underwear', 'shampoo', 'desk', 'piano_bench', 'toaster', 'ceilinglamp', 'window', 'food_peanut_butter', 'clothes_pants', 'clothes_hat', 'basket_for_clothes', 'bed', 'coffe_maker', 'tray', 'doorjamb', 'cards', 'mouse', 'dresser', 'clothes_dress', 'bathtub', 'table', 'dishwasher', 'soap', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I need to get the dirty clothes basket.  Walk to bathroom.  Put all the clothes in washing machine.  Add soap.  Turn on machine.", "actions": ["[WALK] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'woman', 'knifeblock', 'needle', 'cpuscreen', 'desk', 'man', 'keyboard', 'bedroom', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'nightstand', 'oven', 'cup', 'food_food', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'food_salt', 'microwave', 'toilet', 'freezer', 'walllamp', 'blow_dryer', 'bathroom', 'orchid', 'knifeblock', 'drawing', 'desk', 'bedroom', 'bookshelf', 'mat', 'foundation', 'oil', 'pillow', 'dining_room', 'bench', 'oven', 'novel', 'mouthwash', 'television', 'kitchen_counter', 'shoes', 'controller', 'bathtub', 'home_office']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Find the computer.  Turn on the computer.  Open the e-mail program.  Create a new message.  Write the message.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_pizza', 'filing_cabinet', 'toilet', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'folder', 'keyboard', 'faucet', 'iron', 'mousepad', 'pillow', 'dining_room', 'tape', 'cutting_board', 'cup', 'food_food', 'video_game_controller', 'food_rice', 'food_donut', 'rag', 'kettle', 'mouse', 'kitchen_counter', 'controller', 'food_steak', 'sink', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find a radio there and switch it on to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'detergent', 'wallshelf', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'alcohol', 'table_cloth', 'toaster', 'tvstand', 'shower', 'faucet', 'mousepad', 'clothes_hat', 'powersocket', 'oven', 'bench', 'box', 'photoframe', 'novel', 'coffe_maker', 'toilet_paper', 'doorjamb', 'mouse', 'cupboard', 'napkin', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go looking for the cat in the living room where he usually sleeps. Once I find the cat I just pet him for a few minutes.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'door', 'teeth', 'printing_paper']\nTask:Print out papers\nresponse:\n{\n    \"thoughts\": Turn on printer and fill with blank paper. Use computer to click on printing option. Take printed page out of print.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <fax_machine>', '[FIND] <fax_machine>', '[SWITCHON] <fax_machine>', '[FIND] <printing_paper>', '[GRAB] <printing_paper>', '[PUTBACK] <printing_paper> <fax_machine>', '[FIND] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[GRAB] <printing_paper>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'toaster', 'plate', 'faucet', 'stereo', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'coffe_maker', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'bathtub', 'dishwasher', 'clothes_skirt', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Take a dish and put it in the dishwasher to be washed", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'sheets', 'ground_coffee', 'walllamp', 'trashcan', 'drawing', 'orchid', 'needle', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'toaster', 'ceilinglamp', 'couch', 'pillow', 'wall', 'nightstand', 'check', 'food_food', 'television', 'food_onion', 'doorjamb', 'pencil', 'door', 'home_office', 'table', 'light', 'coffee']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Walk to the sofa and sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'knifeblock', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into my living room, find the remote and sit on my sofa. I look at the remote and switch on the television. I look at the Television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'jelly', 'drawing', 'floor', 'bedroom', 'table_cloth', 'stovefan', 'mat', 'plate', 'window', 'couch', 'wall', 'nightstand', 'wall_clock', 'cutting_board', 'bed', 'novel', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'pencil', 'cd', 'clothes_dress', 'bathroom_cabinet', 'table', 'light', 'vase', 'laser_pointer']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I would go lie down in my bed and open the book and start reading.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'walllamp', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'pencil', 'door', 'light', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I push my body to be near to the table in front of me, without stand up. I move my right hand to near the TV remote control on the table. I take the TV remote control. I look the number on the remote control. I touch number 3 button. I put back the remote control to the table. I put my body back to the sofa", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[GRAB] <remote_control>", "[TOUCH] <remote_control>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'detergent', 'food_butter', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'bedroom', 'button', 'table_cloth', 'stovefan', 'mat', 'brush', 'stereo', 'picture', 'mousepad', 'couch', 'clothes_pants', 'food_jam', 'powersocket', 'oven', 'newspaper', 'cards', 'shoes', 'controller', 'clothes_skirt', 'board_game']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go into the living room and turn on the radio.  Then, I select the station I want, which is usually a preset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[FIND] <button>', '[PUSH] <button>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'bathroom', 'detergent', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'bookshelf', 'tvstand', 'faucet', 'mousepad', 'couch', 'pillow', 'wall', 'photoframe', 'check', 'food_food', 'coffe_maker', 'doorjamb', 'dresser', 'door', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'light', 'home_office']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'mop', 'chair', 'freezer', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'toaster', 'shower', 'window', 'food_orange', 'wall', 'oven', 'cup', 'photoframe', 'food_food', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'spoon', 'table', 'light', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'dvd_player', 'character', 'bookshelf', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'laptop', 'couch', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'lighter', 'dresser', 'sink', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'dog', 'chair', 'bathroom', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'mat', 'ceilinglamp', 'shower', 'candle', 'wall', 'bed', 'clothes_gloves', 'food_food', 'water', 'food_rice', 'wooden_spoon', 'curtain', 'console', 'dishrack', 'pajamas', 'light', 'clothes_skirt', 'soap', 'dirt']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['pot', 'tea', 'microwave', 'chair', 'freezer', 'band-aids', 'orchid', 'drawing', 'floor', 'ceiling', 'shampoo', 'alcohol', 'toaster', 'shower', 'mousepad', 'oil', 'couch', 'wall', 'nightstand', 'bench', 'cutting_board', 'tape', 'cup', 'food_donut', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Find groceries. Grab groceries. Walk to refrigerator. Open refrigerator door. Place groceries into refrigerator. Close refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'phone', 'bookshelf', 'table_cloth', 'bookmark', 'toaster', 'mat', 'faucet', 'oven', 'cup', 'photoframe', 'check', 'television', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'door', 'sink', 'light', 'tooth_paste']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'dog', 'chair', 'walllamp', 'detergent', 'knifeblock', 'needle', 'character', 'bookmark', 'mat', 'ceilinglamp', 'faucet', 'pillow', 'powersocket', 'oven', 'tape', 'cutting_board', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'clothes_shirt', 'dirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'desk', 'envelope', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'oven_mitts', 'curtain', 'controller', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room. I cross the room to the couch. I sit down and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'ground_coffee', 'walllamp', 'nail_polish', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'toaster', 'stovefan', 'iron', 'couch', 'form', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'food_food', 'television', 'coffee_filter', 'water', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'hairdryer', 'bathtub', 'dishwasher', 'dirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk into the kitchen. I find the coffee maker and open it. I find the coffee filter, grab it and put it into the coffee maker. I find ground coffee, grab it and put it into the coffee maker. I find some water, then I pour it into the coffee maker. I close the coffee maker, I plug it in and I switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'walllamp', 'ground_coffee', 'trashcan', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'foundation', 'couch', 'microphone', 'wall', 'powersocket', 'oven', 'bed', 'television', 'wooden_spoon', 'doorjamb', 'lighter', 'dresser', 'door', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['spectacles', 'scissors', 'food_ice_cream', 'chair', 'orchid', 'drawing', 'bills', 'knifeblock', 'floor', 'character', 'alcohol', 'mat', 'window', 'couch', 'microphone', 'pillow', 'dining_room', 'wall', 'pasta', 'clothes_gloves', 'check', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'table', 'dishwasher', 'colander', 'coffee']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "walk to entrance hall. find a table. pull table. walk to living rook. push table.", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[PULL] <table>", "[WALK] <home_office>", "[WALK] <table>", "[PUSH] <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mop_bucket', 'ceilinglamp', 'mat', 'electric_shaver', 'mousepad', 'wall', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'mouse', 'pencil', 'controller', 'door', 'table', 'teeth']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to home office ,find computer ,switch on computer and work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'freezer', 'walllamp', 'knifeblock', 'orchid', 'floor', 'ceiling', 'stamp', 'table_cloth', 'toaster', 'ceilinglamp', 'plate', 'shower', 'dish_soap', 'wall', 'nightstand', 'basket_for_clothes', 'keys', 'light_bulb', 'coffe_maker', 'novel', 'television', 'curtain', 'kitchen_counter', 'dishrack', 'cupboard', 'home_office', 'dishwasher', 'scrabble']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the  kitchen. Open the dishwasher. Put plates into the dishwasher. Put dish soap into the dishwasher. Close the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'laptop', 'plate', 'stovefan', 'shower', 'dish_soap', 'wall', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'kettle', 'pencil', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher', 'coffee']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put the dishes on dinner table in the dishwasher and wash them,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[WALK] <table>', '[FIND] <pot>', '[GRAB] <pot>', '[WALK] <dishwasher>', '[PUTBACK] <pot> <dishwasher>', '[CLOSE] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[OPEN] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'chair', 'freezer', 'drawing', 'floor', 'ceiling', 'character', 'toaster', 'plate', 'shower', 'floor_lamp', 'window', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'oven', 'toilet_paper', 'curtain', 'doorjamb', 'towel_rack', 'shoes', 'cd', 'laundry_detergent', 'door', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to wall, find light switch, turn on lights", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'microwave', 'detergent', 'knifeblock', 'floor', 'ceiling', 'character', 'computer', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'wall', 'cup', 'check', 'mouthwash', 'water', 'wine', 'after_shave', 'curtain', 'food_onion', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'conditioner', 'sink', 'napkin', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go in the kitchen and get the glass near the sink. I open the faucet, then I pour water in it, turn off the faucet and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['sponge', 'dry_pasta', 'tablelamp', 'walllamp', 'detergent', 'milk', 'orchid', 'floor', 'ceiling', 'button', 'toaster', 'faucet', 'wall', 'box', 'basket_for_clothes', 'keys', 'coffe_maker', 'food_food', 'curtain', 'food_sugar', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'bathtub', 'sink', 'home_office', 'coffee_cup', 'dishwasher', 'clothes_skirt']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[LOOKAT] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <milk>", "[GRAB] <milk>", "[PUTBACK] <milk> <coffe_maker>", "[FIND] <food_sugar>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[SWITCHON] <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'stovefan', 'wall_clock', 'powersocket', 'bed', 'food_food', 'television', 'wooden_spoon', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'hairdryer', 'home_office', 'light', 'clothes_skirt', 'bathtub', 'bathroom_cabinet']\nTask:Read book\nresponse:\n{\n    \"thoughts\": pick up your book and read it,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'dog', 'bathroom', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'desk', 'shower', 'electric_shaver', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'newspaper', 'bed', 'food_food', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'bathroom_cabinet', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and find the sofa. I sit on the sofa and get comfortable.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'chair', 'walllamp', 'detergent', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'table_cloth', 'mop_bucket', 'ceilinglamp', 'tvstand', 'mat', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'check', 'television', 'tray', 'wooden_spoon', 'kettle', 'controller', 'food_steak', 'sink']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room and sit on the couch, ease into the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'bathroom', 'cat', 'knifeblock', 'drawing', 'wallshelf', 'phone', 'ceiling', 'character', 'keyboard', 'toaster', 'mat', 'tvstand', 'electric_shaver', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'photoframe', 'novel', 'food_food', 'curtain', 'doorjamb', 'dishrack', 'pajamas', 'door', 'bathtub', 'spoon', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to phone, grab phone, pick up phone, put to ear", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'jelly', 'ceilingfan', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'bookshelf', 'computer', 'clothes_jacket', 'piano_bench', 'stovefan', 'toaster', 'tvstand', 'ceilinglamp', 'balanceball', 'stereo', 'brush', 'food_jam', 'clothes_hat', 'nightstand', 'pasta', 'powersocket', 'bench', 'bed', 'oven_mitts', 'tray', 'towel_rack', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to the radio.  I switch on the radio and listen,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'floor', 'phone', 'ceiling', 'keyboard', 'computer', 'toaster', 'mat', 'tvstand', 'couch', 'dustpan', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'newspaper', 'novel', 'doorjamb', 'mouse', 'pencil', 'cupboard', 'bathtub', 'bathroom_cabinet', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the kitchen, I find the light switch, and then i switch on the lights in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'toaster', 'mat', 'ceilinglamp', 'plate', 'stovefan', 'dish_soap', 'broom', 'wall', 'bench', 'food_food', 'coffe_maker', 'tray', 'wooden_spoon', 'towel_rack', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'table', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go the in kitchen then grab the plate on the table. I put the plate and dish soap in the dishwasher. Finally, I turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'chair', 'bathroom', 'knifeblock', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'character', 'bedroom', 'desk', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'homework', 'mousepad', 'form', 'powersocket', 'bench', 'light_bulb', 'check', 'television', 'tray', 'curtain', 'chessboard', 'cupboard', 'home_office']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office, find where the phone is and lift the reciever.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'wallshelf', 'floor', 'phone', 'character', 'bedroom', 'alcohol', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'foundation', 'food_noodles', 'wall', 'nightstand', 'bench', 'oven', 'food_food', 'television', 'curtain', 'rag', 'pencil', 'door', 'clothes_scarf', 'home_office', 'table']\nTask:Play games\nresponse:\n{\n    \"thoughts\": I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <board_game>', '[FIND] <board_game>', '[GRAB] <board_game>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <board_game> <table>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'chair', 'walllamp', 'orchid', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'powersocket', 'oven', 'bench', 'photoframe', 'television', 'curtain', 'food_onion', 'doorjamb', 'towel_rack', 'bathroom_cabinet', 'light', 'home_office']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'dog', 'band-aids', 'walllamp', 'ground_coffee', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'phone', 'food_cake', 'computer', 'toaster', 'mat', 'faucet', 'window', 'couch', 'candle', 'pillow', 'wall', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'table', 'light', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'detergent', 'trashcan', 'orchid', 'wallshelf', 'bathroom_counter', 'ceiling', 'keyboard', 'stovefan', 'tvstand', 'shower', 'window', 'mousepad', 'dining_room', 'wall', 'basket_for_clothes', 'television', 'centerpiece', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'clothes_dress', 'laundry_detergent', 'door', 'sink', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the washing machine. Open the washing machine. Find the basket for clothes. Find the clothes_dress. Grab the clothes_dress. Put the clothes_dress into the washing machine. Find the clothes_shirt. Grab the clothes_shirt. Put the clothes_shirt into the washing machine. Close the washing machine. Switch on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <basket_for_clothes>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'powersocket', 'check', 'television', 'wine', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table', 'clothes_skirt']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": enter office, find computer, turn it on,,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TOUCH] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'freezer', 'detergent', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'bookshelf', 'mail', 'toaster', 'ceilinglamp', 'faucet', 'window', 'pillow', 'dining_room', 'wall', 'powersocket', 'oven', 'novel', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'pencil', 'vacuum_cleaner', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on light in an area", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'trashcan', 'food_carrot', 'drawing', 'food_egg', 'bathroom_counter', 'ceiling', 'keyboard', 'bookshelf', 'mop_bucket', 'tvstand', 'iron', 'oil', 'food_noodles', 'mousepad', 'dustpan', 'towel', 'nightstand', 'cup', 'light_bulb', 'razor', 'food_donut', 'rag', 'shaving_cream', 'towel_rack', 'pencil', 'kitchen_counter', 'bag', 'bathroom_cabinet', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'knife', 'knifeblock', 'wallshelf', 'clothes_underwear', 'kitchen_cabinet', 'floor', 'ceiling', 'bedroom', 'desk', 'bookshelf', 'toaster', 'stovefan', 'faucet', 'dustpan', 'wall', 'oven', 'food_food', 'food_donut', 'tray', 'after_shave', 'pencil', 'kitchen_counter', 'cupboard', 'food_steak', 'sink']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Open fridge. Take food from fridge. Close fridge. Put food on kitchen counter. Prepare food. Turn on stove. Open kitchen cabinet. Take pot. Close cabinet. Put food in pot. Put pot on stove. Cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <kitchen_counter>", "[PUTBACK] <food_food> <kitchen_counter>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_pants', 'form', 'pillow', 'towel', 'wall', 'powersocket', 'bench', 'newspaper', 'television', 'curtain', 'doorjamb', 'mouse', 'hair', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'freezer', 'band-aids', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'stamp', 'desk', 'stovefan', 'shoe-shine_kit', 'couch', 'pillow', 'wall', 'bench', 'coffe_maker', 'toothbrush_holder', 'television', 'curtain', 'doorjamb', 'shaving_cream', 'cupboard', 'door', 'sink', 'bathtub', 'instrument_guitar', 'light', 'home_office', 'clothes_skirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to room. I enter room. I find where the lightswitch is. I turn on the lights with the lightswitch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[LOOKAT] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'bathroom', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'toaster', 'stovefan', 'ceilinglamp', 'picture', 'tooth_paste', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'check', 'television', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'home_office', 'bathroom_cabinet', 'board_game']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'measuring_cup', 'microwave', 'freezer', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'knifeblock', 'phone', 'bedroom', 'desk', 'button', 'stereo', 'faucet', 'picture', 'pillow', 'dining_room', 'nightstand', 'wall', 'bed', 'check', 'toothbrush_holder', 'tray', 'dresser', 'sink', 'home_office', 'table', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the radio; switch on the radio and listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'detergent', 'ceilingfan', 'food_carrot', 'milk', 'drawing', 'knifeblock', 'food_egg', 'floor', 'phone', 'desk', 'bookshelf', 'envelope', 'ceilinglamp', 'window', 'mousepad', 'couch', 'wall', 'food_cheese', 'novel', 'television', 'food_rice', 'after_shave', 'towel_rack', 'kitchen_counter', 'comb', 'table', 'dishwasher', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and find the sofa. I sit on the sofa and get comfortable.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'blender', 'freezer', 'detergent', 'food_carrot', 'food_egg', 'wallshelf', 'floor', 'ceiling', 'food_cake', 'computer', 'ceilinglamp', 'stereo', 'shower', 'foundation', 'faucet', 'oil', 'broom', 'nightstand', 'oven', 'powersocket', 'newspaper', 'coffe_maker', 'television', 'cupboard', 'bathtub', 'napkin', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the radio. Turn on the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'chair', 'band-aids', 'bathroom_counter', 'character', 'desk', 'keyboard', 'purse', 'mat', 'tvstand', 'faucet', 'stereo', 'shower', 'oven', 'cup', 'clothes_gloves', 'novel', 'television', 'oven_mitts', 'wooden_spoon', 'curtain', 'pencil', 'kitchen_counter', 'paper_towel', 'coffee', 'cupboard', 'laundry_detergent', 'dishwasher', 'hands_both']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walking towards Toilet room, And dispose of human waste. Cleaning your hands with tissue paper or water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <paper_towel>', '[FIND] <paper_towel>', '[TURNTO] <paper_towel>', '[LOOKAT] <paper_towel>', '[GRAB] <paper_towel>', '[FIND] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'food_oatmeal', 'chair', 'freezer', 'bathroom', 'food_carrot', 'drawing', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'toaster', 'window', 'clothes_hat', 'dining_room', 'food_cereal', 'wall', 'food_vegetable', 'food_food', 'television', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'pajamas']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_vegetable> <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTIN] <food_carrot> <freezer>", "[FIND] <food_cereal>", "[GRAB] <food_cereal>", "[PUTIN] <food_cereal> <freezer>", "[FIND] <food_oatmeal>", "[GRAB] <food_oatmeal>", "[PUTIN] <food_oatmeal> <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[FIND] <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'jelly', 'walllamp', 'bathroom', 'knife', 'wallshelf', 'floor', 'ceiling', 'bookmark', 'ceilinglamp', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'television', 'toothbrush_holder', 'wooden_spoon', 'kettle', 'shoes', 'cupboard', 'conditioner', 'clothes_dress', 'door', 'napkin', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Pick up a book and open to the last page you were on. Read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <bookmark>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'stereo', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to living room, sit on couch, get remote, switch on tv, watch tv", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'mat', 'shower', 'mousepad', 'wall', 'powersocket', 'oven', 'bench', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'door', 'table', 'light', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go in my office, sit at my desk turn on the computer, then type on my keyboard and open my web browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'tablelamp', 'juice', 'food_salt', 'microwave', 'chair', 'freezer', 'detergent', 'trashcan', 'toothbrush', 'floor', 'ceiling', 'bookshelf', 'toaster', 'shower', 'stereo', 'mousepad', 'clothes_pants', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'food_food', 'check', 'food_donut', 'kitchen_counter', 'cupboard', 'door', 'home_office']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Go kitchen and make something for eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'food_egg', 'bathroom_counter', 'folder', 'bedroom', 'computer', 'table_cloth', 'piano_bench', 'stovefan', 'tvstand', 'faucet', 'foundation', 'homework', 'couch', 'nightstand', 'powersocket', 'oven', 'basket_for_clothes', 'cup', 'keys', 'food_food', 'mouthwash', 'after_shave', 'doorjamb', 'shoes', 'controller', 'maindoor', 'bag', 'dirt', 'coffee']\nTask:Get ready to leave\nresponse:\n{\n    \"thoughts\": Get your jacket and shoes on, locate the keys off the side table, walk out the front door and secure building with keys when you leave,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>', '[FIND] <keys>', '[GRAB] <keys>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_bread', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'shower', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'basket_for_clothes', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'console', 'dresser', 'hair', 'door', 'bathroom_cabinet', 'table', 'light', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I go to the living room and sit in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sheets', 'filing_cabinet', 'food_salt', 'toilet', 'band-aids', 'walllamp', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'phone', 'computer', 'mop_bucket', 'stovefan', 'tvstand', 'deck_of_cards', 'closetdrawer', 'newspaper', 'coffe_maker', 'doorjamb', 'dresser', 'cupboard', 'door', 'food_steak', 'bag', 'sink', 'table', 'dishwasher', 'bathroom_cabinet']\nTask:Put umbrella away\nresponse:\n{\n    \"thoughts\": Place the umbrella in the umbrella bucket point first.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mop_bucket>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'faucet', 'couch', 'food_jam', 'pillow', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "you go into the living room, sit on a chair/sofa, grab remote control, turn on tv, change channels, then watch somethign fun.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'food_salt', 'walllamp', 'bathroom', 'trashcan', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'tvstand', 'shower', 'faucet', 'towel', 'wall', 'cup', 'keys', 'food_food', 'television', 'water', 'curtain', 'lighter', 'kitchen_counter', 'shoes', 'door', 'table', 'light', 'soap', 'teeth']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'jelly', 'freezer', 'bathroom', 'drawing', 'orchid', 'character', 'folder', 'keyboard', 'mop_bucket', 'stovefan', 'faucet', 'food_noodles', 'couch', 'clothes_hat', 'towel', 'nightstand', 'powersocket', 'rag', 'towel_rack', 'pencil', 'kitchen_counter', 'cupboard', 'pajamas', 'bathroom_cabinet', 'table', 'dishwasher', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and get the soap and wash my both hand using the soap and rinse my hands and use towel.", "actions": ["[WALK] <dining_room>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[FIND] <towel>", "[RINSE] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'microwave', 'toilet', 'walllamp', 'drawing', 'bills', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'television', 'tray', 'curtain', 'towel_rack', 'pencil', 'kitchen_counter', 'bathtub', 'spoon', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'chair', 'walllamp', 'blow_dryer', 'knifeblock', 'drawing', 'orchid', 'kitchen_cabinet', 'floor', 'ceiling', 'food_cake', 'mat', 'faucet', 'mousepad', 'couch', 'wall', 'oven', 'bench', 'toilet_paper', 'television', 'water', 'drinking_glass', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen to get a glass from the cabinet. I turn on the faucet and fill up the glass. I take a drink.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <drinking_glass> <sink>", "[FIND] <water>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'oven', 'bench', 'shelf', 'check', 'television', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'light', 'table', 'vase']\nTask:Research\nresponse:\n{\n    \"thoughts\": I read different articles online to gain a better understanding of a particular business, industry, method, terminology, etc.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'foundation', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'cards', 'door', 'table', 'light', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room and sit on the sofa to watch tv", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'tablelamp', 'blender', 'microwave', 'toilet', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'shampoo', 'character', 'desk', 'keyboard', 'computer', 'mousepad', 'wall', 'bench', 'powersocket', 'photoframe', 'television', 'tray', 'mouse', 'dresser', 'controller', 'maindoor', 'home_office', 'clothes_skirt']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to bathroom, find toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'microwave', 'freezer', 'trashcan', 'bathroom_counter', 'ceiling', 'desk', 'stovefan', 'stereo', 'shower', 'mousepad', 'dustpan', 'clothes_pants', 'dining_room', 'powersocket', 'bench', 'bed', 'toilet_paper', 'food_food', 'wine', 'mouse', 'pencil', 'kitchen_counter', 'cd', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room ,find radio ,switch on radio ,listen to music 4", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'computer', 'tvstand', 'faucet', 'stereo', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'wall', 'powersocket', 'shelf', 'food_food', 'oven_mitts', 'curtain', 'doorjamb', 'coffee_pot', 'mouse', 'vacuum_cleaner', 'dresser', 'light', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": approach the light switch, raise hand to the switch, flip the switch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'freezer', 'bathroom', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'bookshelf', 'shower', 'homework', 'wall', 'powersocket', 'box', 'bed', 'coffe_maker', 'television', 'after_shave', 'doorjamb', 'towel_rack', 'mouse', 'cards', 'cupboard', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into living room. Find wall. Find light button. Switch on light button.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'clothes_jacket', 'table_cloth', 'mat', 'floor_lamp', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'light_bulb', 'novel', 'mouthwash', 'television', 'ice', 'curtain', 'after_shave', 'doorjamb', 'console', 'kettle', 'conditioner', 'bathtub', 'sink', 'table', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['freezer', 'drawing', 'orchid', 'knifeblock', 'hanger', 'floor', 'ceiling', 'dvd_player', 'stamp', 'bedroom', 'desk', 'keyboard', 'toaster', 'faucet', 'shower', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'bed', 'toothbrush_holder', 'doorjamb', 'towel_rack', 'door', 'bathtub', 'home_office', 'table', 'light', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music", "actions": ["[WALK] <home_office>", "[WALK] <dvd_player>", "[FIND] <dvd_player>", "[TURNTO] <dvd_player>", "[LOOKAT] <dvd_player>", "[SWITCHON] <dvd_player>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'television', 'oven_mitts', 'tray', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <chair>', '[SIT] <chair>', '[TURNTO] <television>', '[WATCH] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'character', 'bookshelf', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'powersocket', 'bench', 'oven', 'bed', 'novel', 'television', 'tray', 'doorjamb', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'light', 'table', 'laser_pointer']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "My friend is calling and I want to pick up the phone and chat with him for a little while on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>", "[TYPE] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'plate', 'faucet', 'couch', 'pillow', 'dining_room', 'broom', 'wall', 'powersocket', 'box', 'photoframe', 'novel', 'food_food', 'television', 'food_donut', 'curtain', 'doorjamb', 'bathroom_cabinet']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on couch. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <couch>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_bread', 'food_ice_cream', 'crayon', 'microwave', 'chair', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'faucet', 'tvstand', 'window', 'wall', 'bench', 'powersocket', 'clothes_gloves', 'water', 'tray', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'light', 'instrument_violin']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, hold drinking glass, lift glass, pour water in drinking glass", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[FIND] <water>", "[GRAB] <water>", "[WALK] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'dough', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookmark', 'stovefan', 'brush', 'deck_of_cards', 'faucet', 'window', 'closetdrawer', 'dining_room', 'wall', 'oven', 'cutting_board', 'cup', 'bed', 'coffe_maker', 'food_food', 'tray', 'shaving_cream', 'kitchen_counter', 'chessboard', 'clothes_dress', 'conditioner']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[TURNTO] <cup>', '[LOOKAT] <cup>', '[WALK] <cup>', '[FIND] <freezer>', '[TURNTO] <freezer>', '[LOOKAT] <freezer>', '[WALK] <freezer>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'blender', 'tea', 'microwave', 'walllamp', 'bathroom', 'knifeblock', 'floor', 'desk', 'tea_bag', 'bookshelf', 'computer', 'bookmark', 'toaster', 'faucet', 'stereo', 'dish_soap', 'dining_room', 'wall', 'nightstand', 'bench', 'coffe_maker', 'razor', 'tray', 'kitchen_counter', 'cupboard', 'controller', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen. Look at sink. Grab dish soap. Open dishwasher. Put dish soap in dishwasher. Close dishwasher. Switch on dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[LOOKAT] <sink>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink']\nTask:Use computer\nresponse:\n{\n    \"thoughts\": Browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'food_butter', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'mop_bucket', 'ceilinglamp', 'tvstand', 'iron', 'window', 'mousepad', 'microphone', 'towel', 'dining_room', 'bowl', 'wall', 'powersocket', 'bench', 'cup', 'novel', 'television', 'coin', 'rag', 'doorjamb', 'headset', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk in to the living room. I find the light switch. I turn on the lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'walllamp', 'cat', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'bowl', 'wall', 'powersocket', 'box', 'photoframe', 'television', 'doorjamb', 'dresser', 'conditioner', 'door', 'coffee']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": We sit all around the television and watch a program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'faucet', 'stereo', 'shower', 'candle', 'wall', 'oven', 'cup', 'photoframe', 'water', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, walk to cupboard, open cupboard, grab cup, walk to sink, pour water in cup, drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'blow_dryer', 'milk', 'bills', 'orchid', 'floor', 'ceiling', 'character', 'food_cake', 'alcohol', 'stovefan', 'homework', 'closetdrawer', 'pillow', 'broom', 'wall', 'oven', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'kitchen_counter', 'shoes', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'phone', 'ceiling', 'bookshelf', 'computer', 'table_cloth', 'toaster', 'mat', 'picture', 'wall', 'powersocket', 'bench', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'sink', 'bathroom_cabinet', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the dining room and turn on the lights", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'trashcan', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'thread', 'computer', 'piano_bench', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'oven', 'bed', 'novel', 'check', 'television', 'curtain', 'towel_rack', 'dresser', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Put child in bed, sit down,read book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'microwave', 'drawing', 'knife', 'bedroom', 'tea_bag', 'computer', 'brush', 'ceilinglamp', 'plate', 'faucet', 'mousepad', 'oil', 'pillow', 'wall', 'powersocket', 'tape', 'photoframe', 'light_bulb', 'novel', 'check', 'television', 'wooden_spoon', 'curtain', 'tray', 'shaving_cream', 'clothes_scarf', 'sink', 'dishwasher']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go to home office, turn on the computer, open the internet and browse around the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'light', 'table']\nTask:Straighten paintings on wall\nresponse:\n{\n    \"thoughts\": Go up to the painting and tilt it which ever way it is uneven until it looks straight.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <wall>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['band-aids', 'trashcan', 'wallshelf', 'clothes_underwear', 'desk', 'keyboard', 'mail', 'computer', 'tvstand', 'shower', 'window', 'clothes_socks', 'couch', 'food_peanut_butter', 'pillow', 'basket_for_clothes', 'keys', 'coffe_maker', 'television', 'food_donut', 'tray', 'rag', 'headset', 'clothes_scarf', 'bathtub', 'sink', 'light', 'dishwasher', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'console', 'dresser', 'cupboard', 'home_office', 'table', 'clothes_skirt', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'toilet', 'freezer', 'cpuscreen', 'clothes_underwear', 'wallshelf', 'shampoo', 'character', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'clothes_pants', 'dining_room', 'bench', 'powersocket', 'photoframe', 'coffe_maker', 'mouse', 'paper_towel', 'dresser', 'door', 'bathtub', 'bathroom_cabinet', 'light', 'hands_both']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walking towards Toilet room, And dispose of human waste. Cleaning your hands with tissue paper or water.", "actions": ["[WALK] <bathroom>", "[WALK] <paper_towel>", "[FIND] <paper_towel>", "[TURNTO] <paper_towel>", "[LOOKAT] <paper_towel>", "[GRAB] <paper_towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'toilet', 'chair', 'freezer', 'wallshelf', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'toaster', 'ceilinglamp', 'window', 'pillow', 'wall', 'powersocket', 'oven', 'bench', 'food_cheese', 'bed', 'photoframe', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'lighter', 'kitchen_counter', 'cupboard', 'home_office']\nTask:Eat cheese\nresponse:\n{\n    \"thoughts\": Open fridge.  Get cheese.  Eat it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[EAT] <food_cheese>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'computer', 'mat', 'shower', 'mousepad', 'couch', 'form', 'pillow', 'wall', 'wall_clock', 'powersocket', 'bench', 'oven', 'tape', 'bed', 'check', 'doorjamb', 'cd', 'hair', 'door', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I sit on the sofa. I put my legs on the sofa and recline.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'toilet', 'jelly', 'ground_coffee', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'clothes_jacket', 'faucet', 'mousepad', 'wall', 'oven', 'cup', 'photoframe', 'video_game_controller', 'television', 'tray', 'doorjamb', 'mouse', 'conditioner', 'sink', 'bathroom_cabinet', 'home_office', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom and wash hands.,\n    \"actions\": ['[RUN] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <sink>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[GRAB] <soap>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'toaster', 'mat', 'tvstand', 'shower', 'shoe-shine_kit', 'window', 'mousepad', 'clothes_socks', 'couch', 'food_noodles', 'pillow', 'wall', 'bed', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'home_office', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into the bedroom and walk to the light switch. Flip the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'cat', 'trashcan', 'milk', 'hanger', 'bathroom_counter', 'desk', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'dining_room', 'nightstand', 'wall', 'oven', 'cup', 'bed', 'coffe_maker', 'tray', 'curtain', 'chessboard', 'shoes', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Petting a cat may sound simple.Letting the Cat Come to You.Extend a hand or finger and allow the cat a chance to touch her nose to you. When a cat bumps her head into your hand, it's a signal she wants attention.Focusing on Areas with Scent Glands. Feed with food.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <milk>', '[FIND] <milk>', '[GRAB] <milk>', '[POUR] <milk> <cup>', '[TURNTO] <cat>', '[POINTAT] <cat>']\n}\n", "input": "\nScene_Objects:['walllamp', 'trashcan', 'phone', 'ceiling', 'character', 'desk', 'bookmark', 'toaster', 'stovefan', 'shoe-shine_kit', 'shower', 'window', 'wall', 'pasta', 'oven', 'powersocket', 'clothes_gloves', 'novel', 'coffe_maker', 'check', 'coin', 'curtain', 'doorjamb', 'cards', 'towel_rack', 'kitchen_counter', 'chessboard', 'napkin', 'bathroom_cabinet', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room ,find electrical outlet , switch on eletrical outlet", "actions": ["[WALK] <home_office>", "[WALK] <electrical_outlet>", "[FIND] <electrical_outlet>", "[SWITCHON] <electrical_outlet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'detergent', 'food_butter', 'trashcan', 'hanger', 'floor', 'phone', 'ceiling', 'bedroom', 'computer', 'button', 'mat', 'laptop', 'plate', 'window', 'wall', 'nightstand', 'bed', 'check', 'curtain', 'doorjamb', 'dresser', 'cupboard', 'controller', 'door', 'light', 'dishwasher', 'laser_pointer', 'shoe_rack']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": first i will go to home office and switch on the lights and answer the call after then i will put back it and switch off the lights and close the door,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <phone>', '[GRAB] <phone>', '[PUTOBJBACK] <phone>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'mop', 'chair', 'freezer', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'toaster', 'shower', 'window', 'food_orange', 'wall', 'oven', 'cup', 'photoframe', 'food_food', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'spoon', 'table', 'light', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'food_salt', 'microwave', 'walllamp', 'dough', 'hanger', 'ceiling', 'phone', 'character', 'keyboard', 'bookmark', 'toaster', 'mat', 'tvstand', 'stereo', 'window', 'oil', 'food_peanut_butter', 'tape', 'check', 'video_game_controller', 'razor', 'oven_mitts', 'food_rice', 'curtain', 'kitchen_counter', 'hairbrush', 'food_steak', 'bathroom_cabinet', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into the bedroom. I walk to stereo. I turn on stereo.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'bathroom', 'detergent', 'food_carrot', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'thread', 'stereo', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'cup', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'hair', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room, turn to the sofa. Sit on the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'ground_coffee', 'walllamp', 'knife', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bedroom', 'button', 'stovefan', 'mat', 'shower', 'clothes_socks', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'bed', 'light_bulb', 'check', 'video_game_controller', 'mouthwash', 'food_onion', 'vacuum_cleaner', 'pajamas', 'cupboard', 'hair', 'instrument_guitar']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go to bedroom, put pajamas on, lie in bed, sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <pajamas>', '[FIND] <pajamas>', '[GRAB] <pajamas>', '[PUTON] <pajamas>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['pot', 'blender', 'toilet', 'bathroom', 'detergent', 'trashcan', 'drawing', 'knife', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'button', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'bench', 'food_cheese', 'bed', 'television', 'doorjamb', 'mouse', 'pencil', 'door', 'bathtub', 'napkin', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the bedroom and use the button on the wall, next to the door to turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'crayon', 'microwave', 'toilet', 'detergent', 'milk', 'folder', 'bedroom', 'toaster', 'stovefan', 'balanceball', 'stereo', 'window', 'mousepad', 'food_noodles', 'dining_room', 'nightstand', 'box', 'newspaper', 'tray', 'lighter', 'headset', 'controller', 'door', 'maindoor', 'instrument_guitar', 'table', 'soap']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the entrance hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'button', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'faucet', 'mousepad', 'dining_room', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'rag', 'doorjamb', 'kitchen_counter', 'light', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door']\nTask:Watch youtube\nresponse:\n{\n    \"thoughts\": Go to youtube.Com watch videos,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'walllamp', 'trashcan', 'drawing', 'orchid', 'knifeblock', 'wallshelf', 'food_egg', 'floor', 'ceiling', 'character', 'stovefan', 'tvstand', 'couch', 'candle', 'pillow', 'dining_room', 'wall', 'novel', 'check', 'video_game_controller', 'coffe_maker', 'doorjamb', 'towel_rack', 'cupboard', 'controller', 'door', 'table', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on couch. Grab book. Open book. Read book.", "actions": ["[WALK] <couch>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'floor', 'character', 'bookshelf', 'computer', 'toaster', 'mat', 'plate', 'window', 'food_noodles', 'food_orange', 'wall', 'oven', 'photoframe', 'keys', 'bed', 'coffe_maker', 'toothbrush_holder', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'clothes_dress', 'napkin', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['freezer', 'floor', 'ceiling', 'character', 'computer', 'brush', 'mat', 'tvstand', 'iron', 'foundation', 'mousepad', 'couch', 'food_kiwi', 'oven', 'powersocket', 'bench', 'toilet_paper', 'coin', 'food_rice', 'after_shave', 'doorjamb', 'kitchen_counter', 'dresser', 'conditioner', 'bathtub', 'home_office', 'dishwasher', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'food_pizza', 'juice', 'board_game', 'dog', 'food_butter', 'orchid', 'drawing', 'bathroom_counter', 'bookshelf', 'stovefan', 'mat', 'laptop', 'tvstand', 'picture', 'window', 'couch', 'nightstand', 'wall_clock', 'toy', 'clothes_gloves', 'toilet_paper', 'television', 'ice', 'bathroom_cabinet', 'table', 'light', 'vase', 'instrument_violin']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <home_office>', '[DROP] <plate>', '[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <home_office>', '[DROP] <cup>', '[WALK] <dining_room>', '[WALK] <fork>', '[FIND] <fork>', '[GRAB] <fork>', '[WALK] <home_office>', '[DROP] <fork>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'microwave', 'chair', 'freezer', 'ground_coffee', 'detergent', 'bills', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'alcohol', 'table_cloth', 'toaster', 'stovefan', 'wall', 'oven', 'bench', 'food_food', 'wooden_spoon', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'sink', 'home_office', 'light', 'soap', 'scrabble']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen counter, grab a grocery, walk to fridge, open fridge, put the grocery into fridge, and close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'dresser', 'hairbrush', 'clothes_dress', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'plate', 'mousepad', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'wine', 'curtain', 'doorjamb', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'cd_player', 'blow_dryer', 'bathroom', 'detergent', 'orchid', 'bills', 'bathroom_counter', 'phone', 'bedroom', 'remote_control', 'thread', 'button', 'mat', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'broom', 'wall_clock', 'powersocket', 'bench', 'box', 'coffe_maker', 'razor', 'towel_rack', 'vacuum_cleaner', 'bathtub', 'dishwasher', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[PLUGIN] <cd_player>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['chair', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'ceilinglamp', 'shower', 'window', 'homework', 'mousepad', 'pillow', 'powersocket', 'photoframe', 'coffe_maker', 'doorjamb', 'kettle', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'clothes_dress', 'door', 'bag', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to bedroom and grab dirty pants and take to the washing machine in entrance hall. Put clothes in washing machine and then put in laundry detergent.", "actions": ["[WALK] <bedroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'freezer', 'walllamp', 'detergent', 'cat', 'knife', 'drawing', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'phone', 'ceiling', 'character', 'faucet', 'window', 'foundation', 'closetdrawer', 'food_orange', 'dining_room', 'coffe_maker', 'check', 'television', 'food_food', 'tray', 'curtain', 'mouse', 'pajamas', 'door', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'food_cake', 'button', 'toaster', 'tvstand', 'faucet', 'foundation', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'food_cheese', 'bed', 'toilet_paper', 'coffe_maker', 'doorjamb', 'kettle', 'dresser', 'clothes_dress', 'door', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I'm tired and want to relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'novel', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I walk to the home office. I sit on the chair. I switch on the computer. I look at my monitor. I open the browsing application by click the application using mouse. I type the book that I need to searching for (using keyboard). I found the book that I need. I read the book. I switch on the printer. I print the information that I get. I read carefully the book information that I got. I switch off the printer. I switch of the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <fax_machine>', '[SWITCHON] <fax_machine>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[SWITCHOFF] <fax_machine>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'freezer', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'table_cloth', 'ceilinglamp', 'faucet', 'couch', 'clothes_hat', 'pillow', 'wall', 'oven', 'bed', 'photoframe', 'television', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I would walk into the living room. Then sit on the couch and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'microwave', 'cd_player', 'freezer', 'band-aids', 'bathroom', 'food_carrot', 'orchid', 'music_stand', 'phone', 'bookshelf', 'thread', 'mail', 'stereo', 'homework', 'mousepad', 'closetdrawer', 'clothes_hat', 'form', 'broom', 'newspaper', 'photoframe', 'check', 'food_rice', 'food_donut', 'dresser', 'sink', 'napkin', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to the living and turn to the music stand. Switch on the CD player to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <music_stand>', '[FIND] <music_stand>', '[TURNTO] <music_stand>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'piano_bench', 'ceilinglamp', 'tvstand', 'stereo', 'couch', 'pillow', 'dining_room', 'wall', 'bed', 'food_donut', 'tray', 'curtain', 'doorjamb', 'coffee_pot', 'dresser', 'pajamas', 'door', 'home_office', 'table', 'light', 'instrument_violin']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'freezer', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'envelope', 'computer', 'stereo', 'mousepad', 'closetdrawer', 'oil', 'clothes_pants', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'coin', 'rag', 'kettle', 'mouse', 'chessboard', 'dresser', 'light', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Find dresser, grab pants, grab shirt, put on pants, put on shirt,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'cd_player', 'chair', 'knifeblock', 'music_stand', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'envelope', 'food_bacon', 'mop_bucket', 'stovefan', 'ceilinglamp', 'shower', 'nightstand', 'bed', 'photoframe', 'coffe_maker', 'tray', 'curtain', 'lighter', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to the living and turn to the music stand. Switch on the CD player to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <music_stand>", "[FIND] <music_stand>", "[TURNTO] <music_stand>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'walllamp', 'trashcan', 'dough', 'toothbrush', 'floor', 'dvd_player', 'bedroom', 'stovefan', 'mat', 'tvstand', 'window', 'food_noodles', 'dining_room', 'nightstand', 'powersocket', 'oven', 'cutting_board', 'cup', 'light_bulb', 'coffe_maker', 'mouthwash', 'oven_mitts', 'doorjamb', 'diary', 'bag', 'home_office', 'scrabble']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": go to living room, find phone, pick up phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['mop', 'microwave', 'freezer', 'ground_coffee', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'piano_bench', 'shoe-shine_kit', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'tape', 'food_food', 'doorjamb', 'headset', 'cupboard', 'door', 'home_office', 'table', 'light', 'bathroom_cabinet', 'board_game']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I get home from work and walk into the living room. I see the couch and go to sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toothbrush', 'knife', 'drawing', 'desk', 'bookshelf', 'purse', 'computer', 'table_cloth', 'brush', 'tvstand', 'shoe-shine_kit', 'picture', 'faucet', 'wall', 'bench', 'cutting_board', 'shelf', 'light_bulb', 'razor', 'after_shave', 'doorjamb', 'diary', 'kitchen_counter', 'dishrack', 'dresser', 'home_office', 'table', 'vase', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the sink. Take dishes out of sink. Open dishwasher and place dishes in dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'freezer', 'trashcan', 'orchid', 'drawing', 'bills', 'wallshelf', 'cpuscreen', 'knife', 'ceiling', 'character', 'stamp', 'keyboard', 'computer', 'ceilinglamp', 'couch', 'pillow', 'pasta', 'bench', 'oven', 'keys', 'novel', 'mouthwash', 'ice', 'tray', 'towel_rack', 'mouse', 'cupboard', 'bathtub']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the TV remote. I change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'fork', 'tablelamp', 'food_salt', 'microwave', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'nightstand', 'bed', 'novel', 'television', 'curtain', 'lighter', 'mouse', 'door', 'table', 'dishwasher']\nTask:Admire art\nresponse:\n{\n    \"thoughts\": Looking at painting, art,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'microwave', 'chair', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'desk', 'keyboard', 'window', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'bed', 'newspaper', 'photoframe', 'food_rice', 'doorjamb', 'mouse', 'cupboard', 'door', 'hairdryer', 'sink', 'table', 'light', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room and sit on the couch, ease into the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'bathroom', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'mousepad', 'dining_room', 'wall', 'light_bulb', 'check', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'clothes_dress', 'door', 'table', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'sponge', 'food_salt', 'tea', 'bathroom', 'detergent', 'knifeblock', 'wallshelf', 'clothes_underwear', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'toaster', 'faucet', 'window', 'clothes_socks', 'clothes_pants', 'pillow', 'bench', 'basket_for_clothes', 'coffe_maker', 'towel_rack', 'clothes_dress', 'laundry_detergent', 'table', 'clothes_skirt', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom, take the dirt clothes and put into the washing machine, then put the soap, set the cleaning program and switch on the machine", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_skirt>", "[GRAB] <clothes_skirt>", "[PUTBACK] <clothes_skirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[CLOSE] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'detergent', 'orchid', 'bathroom_counter', 'character', 'keyboard', 'remote_control', 'bookshelf', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'dustpan', 'nightstand', 'bench', 'tape', 'cup', 'bed', 'coffee_table', 'wooden_spoon', 'doorjamb', 'towel_rack', 'pencil', 'kitchen_counter', 'cd', 'dresser', 'hair']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['dog', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'bathtub', 'sink', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk into living room and go to sofa and turn on tv", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[WALK] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[WALK] <couch>", "[LIE] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[DROP] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'bed', 'television', 'curtain', 'pencil', 'dresser', 'clothes_dress', 'conditioner', 'table', 'board_game', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Change the channel on the TV in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'toilet', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'ceilinglamp', 'tvstand', 'stereo', 'faucet', 'mousepad', 'dining_room', 'wall', 'powersocket', 'food_food', 'food_donut', 'towel_rack', 'mouse', 'shoes', 'dishrack', 'sink', 'spoon', 'light', 'coffee']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to bathroom, find toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'walllamp', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'pencil', 'door', 'light', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I push my body to be near to the table in front of me, without stand up. I move my right hand to near the TV remote control on the table. I take the TV remote control. I look the number on the remote control. I touch number 3 button. I put back the remote control to the table. I put my body back to the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[GRAB] <remote_control>', '[TOUCH] <remote_control>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'mop', 'chair', 'freezer', 'walllamp', 'knifeblock', 'clothes_underwear', 'floor', 'ceiling', 'character', 'stovefan', 'toaster', 'plate', 'ceilinglamp', 'faucet', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'food_cheese', 'food_food', 'tray', 'curtain', 'kitchen_counter', 'sauce_pan', 'cupboard', 'sink', 'spoon', 'coffee']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <spoon>", "[GRAB] <spoon>", "[FIND] <sauce_pan>", "[PUTBACK] <spoon> <sauce_pan>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <plate>", "[PUTBACK] <pot> <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'mop', 'microwave', 'trashcan', 'food_egg', 'keyboard', 'mop_bucket', 'bookmark', 'mat', 'tvstand', 'mousepad', 'tooth_paste', 'food_noodles', 'food_jam', 'form', 'pillow', 'dining_room', 'food_kiwi', 'toy', 'keys', 'check', 'mouthwash', 'coffee_filter', 'shaving_cream', 'comb', 'maindoor', 'light', 'board_game', 'teeth', 'creditcard']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to living room, walk to the phone reach for the phone and pick up the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TOUCH] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['dog', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'coin', 'curtain', 'doorjamb', 'coffee_pot', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'jelly', 'freezer', 'bathroom', 'trashcan', 'milk', 'toothbrush', 'drawing', 'bathroom_counter', 'floor', 'desk', 'bookshelf', 'alcohol', 'table_cloth', 'faucet', 'shower', 'form', 'powersocket', 'bench', 'bed', 'food_food', 'kitchen_counter', 'headset', 'vacuum_cleaner', 'cupboard', 'door', 'clothes_scarf', 'food_steak', 'soap']\nTask:Grab things\nresponse:\n{\n    \"thoughts\": The robot will need some kind of hands or grabbing device to transport objects.,\n    \"actions\": ['[WALK] <novel>', '[TURNTO] <novel>', '[GRAB] <novel>']\n}\n", "input": "\nScene_Objects:['beer', 'mop', 'walllamp', 'knife', 'wallshelf', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'stamp', 'keyboard', 'toaster', 'plate', 'shower', 'faucet', 'dish_soap', 'wall', 'cup', 'bed', 'coffe_maker', 'video_game_controller', 'check', 'towel_rack', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'conditioner', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>", "[GRAB] <plate>", "[WIPE] <plate>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[PUTOBJBACK] <dishrack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'freezer', 'detergent', 'hanger', 'bathroom_counter', 'shampoo', 'ceiling', 'desk', 'bedroom', 'computer', 'stovefan', 'toaster', 'tvstand', 'shower', 'homework', 'closetdrawer', 'bench', 'oven', 'toy', 'cup', 'bed', 'photoframe', 'toilet_paper', 'coffe_maker', 'oven_mitts', 'pencil', 'dresser', 'bag', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'bathroom', 'detergent', 'trashcan', 'keyboard', 'button', 'tvstand', 'stereo', 'food_jam', 'dining_room', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'newspaper', 'bed', 'coffe_maker', 'water', 'ice', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'table', 'dishwasher', 'tooth_paste', 'laser_pointer']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, look at the water glass, grab that glass, find the water cane, open the water cane tap, drink water", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[TURNTO] <water_glass>", "[LOOKAT] <water_glass>", "[GRAB] <water_glass>", "[FIND] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'milk', 'floor', 'ceiling', 'shampoo', 'purse', 'ceilinglamp', 'mat', 'faucet', 'oil', 'wall', 'pasta', 'tape', 'novel', 'food_food', 'food_donut', 'wooden_spoon', 'mouse', 'kitchen_counter', 'door', 'home_office', 'bathroom_cabinet', 'vase', 'soap', 'instrument_violin', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'chair', 'freezer', 'ground_coffee', 'bathroom', 'food_carrot', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'mail', 'button', 'computer', 'toaster', 'tvstand', 'window', 'wall', 'nightstand', 'cup', 'check', 'television', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "In the kitchen where I left the bags of groceries, I grab cheese from the grocery bag, I walk over to the refrigerator, open it up, neatly put cheese in the fridge, Close the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'food_bread', 'dry_pasta', 'tablelamp', 'filing_cabinet', 'measuring_cup', 'microwave', 'toilet', 'chair', 'walllamp', 'floor', 'bookshelf', 'toaster', 'mat', 'faucet', 'shower', 'closetdrawer', 'bowl', 'dining_room', 'wall', 'newspaper', 'coffe_maker', 'oven_mitts', 'diary', 'kitchen_counter', 'cupboard', 'clothes_dress', 'sink', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to Kitchen. Take the dishes from the sink. Apply dish washer, scrub and then wash with water.,\n    \"actions\": ['[WALK] <sink>', '[FIND] <bowl>', '[GRAB] <bowl>', '[FIND] <dishwasher>', '[SCRUB] <dishwasher>', '[RINSE] <bowl>', '[OPEN] <dishwasher>', '[WASH] <bowl>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'toy', 'novel', 'coffe_maker', 'television', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'home_office', 'table', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the living room and find the book you want to read. Take it and then go to the couch. Sit on it, open the book and start reading.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'beer', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'food_cake', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'photoframe', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'door', 'bathtub', 'bathroom_cabinet', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Sit on sofa. Lie down on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['water_glass', 'tablelamp', 'chair', 'freezer', 'detergent', 'food_carrot', 'knife', 'cpuscreen', 'kitchen_cabinet', 'character', 'bedroom', 'mail', 'stovefan', 'laptop', 'form', 'pillow', 'nightstand', 'bench', 'oven', 'cup', 'keys', 'toothbrush_holder', 'water', 'food_rice', 'curtain', 'controller', 'sink', 'bathtub', 'dishwasher', 'dirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a glass of water to drink", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'food_salt', 'toilet', 'bills', 'hanger', 'bathroom_counter', 'floor', 'character', 'stovefan', 'tvstand', 'faucet', 'mousepad', 'towel', 'dining_room', 'wall', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'water', 'tray', 'curtain', 'doorjamb', 'sink', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Walk to sink and turn on the water. Grab the soap and wash the hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <water>', '[TURNTO] <water>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'scissors', 'chair', 'freezer', 'knifeblock', 'clothes_underwear', 'bookshelf', 'mop_bucket', 'stovefan', 'shower', 'window', 'couch', 'dining_room', 'nightstand', 'wall_clock', 'powersocket', 'photoframe', 'keys', 'bed', 'towel_rack', 'cd', 'dresser', 'cupboard', 'sink', 'napkin', 'bathroom_cabinet', 'home_office', 'dishwasher', 'board_game']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat.  Pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'tea', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'trashcan', 'toothbrush', 'orchid', 'needle', 'shampoo', 'tea_bag', 'alcohol', 'mat', 'shower', 'mousepad', 'towel', 'wall', 'cup', 'bed', 'light_bulb', 'coffe_maker', 'wooden_spoon', 'shoes', 'home_office', 'vase', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk into the living room, walk over to phone, grab the phone, put it to your ear .,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'food_ice_cream', 'microwave', 'freezer', 'walllamp', 'nail_polish', 'bathroom', 'knifeblock', 'floor', 'character', 'desk', 'computer', 'toaster', 'stovefan', 'plate', 'mat', 'dish_soap', 'clothes_hat', 'wall', 'photoframe', 'coffe_maker', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'home_office', 'dishwasher', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk into kitchen, find dishwasher, put plates in dishwasher, put soap in dishwasher, turn on dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'thread', 'computer', 'stovefan', 'mat', 'tvstand', 'toaster', 'couch', 'pillow', 'wall', 'photoframe', 'check', 'television', 'doorjamb', 'laundry_detergent', 'conditioner', 'sink', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Do homework\nresponse:\n{\n    \"thoughts\": Sit on the couch in the Living room and pick u some papers,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <check>', '[GRAB] <check>', '[DROP] <check>']\n}\n", "input": "\nScene_Objects:['spectacles', 'chair', 'bills', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'bookshelf', 'brush', 'ceilinglamp', 'tvstand', 'stereo', 'stovefan', 'mat', 'mousepad', 'wall', 'powersocket', 'bed', 'television', 'doorjamb', 'dresser', 'clothes_scarf', 'light', 'dishwasher', 'table', 'colander', 'clothes_shirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the Living Room. Switch on 1 light. Walk to dining room. Switch on 1 light. Walk to Entrance hall. Switch on 1 light.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'freezer', 'bathroom', 'trashcan', 'dough', 'knifeblock', 'orchid', 'floor', 'ceiling', 'clothes_jacket', 'brush', 'ceilinglamp', 'mat', 'balanceball', 'clothes_socks', 'couch', 'microphone', 'form', 'wall', 'food_food', 'check', 'oven_mitts', 'wine', 'mouse', 'cupboard', 'food_steak', 'spoon', 'table', 'tooth_paste']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Go kitchen and make something for eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'thread', 'computer', 'alcohol', 'stovefan', 'ceilinglamp', 'couch', 'pillow', 'food_kiwi', 'nightstand', 'powersocket', 'check', 'food_food', 'pencil', 'dishrack', 'home_office', 'dishwasher']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go find a computer, turn it on, click google, and browse the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'sheets', 'pot', 'blender', 'microwave', 'chair', 'nail_polish', 'detergent', 'milk', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'clothes_socks', 'broom', 'wall', 'food_kiwi', 'powersocket', 'cup', 'photoframe', 'keys', 'newspaper', 'curtain', 'lighter', 'pencil', 'pajamas', 'clothes_dress', 'sink', 'table', 'light']\nTask:Sit in chair\nresponse:\n{\n    \"thoughts\": I sit in a chair to rest and eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'folder', 'bookshelf', 'remote_control', 'button', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'sink', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go the living room, look for my tv remote control, take the control and then sit down on the couch and turn on the TV and push the buttons on the remote control to change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'measuring_cup', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'bench', 'cup', 'photoframe', 'novel', 'check', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'bathroom_cabinet', 'light', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the book case, I select a book, I go back and sit down, then I read.,\n    \"actions\": ['[WALK] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'door', 'table', 'light', 'dishwasher']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go in the living room. Then I find the remote, sit on the couch and turn on the television. Then I use the remote the change the TV channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'freezer', 'bathroom', 'trashcan', 'dough', 'knifeblock', 'orchid', 'floor', 'ceiling', 'clothes_jacket', 'brush', 'ceilinglamp', 'mat', 'balanceball', 'clothes_socks', 'couch', 'microphone', 'form', 'wall', 'food_food', 'check', 'oven_mitts', 'wine', 'mouse', 'cupboard', 'food_steak', 'spoon', 'table', 'tooth_paste']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Go kitchen and make something for eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'juice', 'food_ice_cream', 'bathroom', 'orchid', 'floor', 'ceiling', 'phone', 'shampoo', 'bookshelf', 'tvstand', 'stereo', 'shower', 'couch', 'wall', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'food_food', 'television', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'controller', 'home_office', 'light', 'coffee']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "first i will go to home office and switch on the lights and answer the call after then i will put back it and switch off the lights and close the door", "actions": ["[RUN] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <phone>", "[GRAB] <phone>", "[PUTOBJBACK] <phone>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tea', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'phone', 'keyboard', 'food_cake', 'toaster', 'stovefan', 'plate', 'mat', 'dish_soap', 'oil', 'pillow', 'wall', 'cup', 'photoframe', 'coffe_maker', 'curtain', 'centerpiece', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'sponge', 'tablelamp', 'measuring_cup', 'toilet', 'chair', 'jelly', 'freezer', 'ground_coffee', 'trashcan', 'toothbrush', 'drawing', 'clothes_underwear', 'phone', 'desk', 'keyboard', 'computer', 'mousepad', 'television', 'towel_rack', 'cd', 'hairbrush', 'pajamas', 'dresser', 'controller', 'door', 'bathtub', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "before going to bed or when wake up earily morning brush teeth", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'walllamp', 'milk', 'floor', 'keyboard', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'faucet', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'toy', 'basket_for_clothes', 'coffe_maker', 'video_game_controller', 'tray', 'doorjamb', 'kettle', 'dresser', 'cupboard', 'clothes_dress', 'sink', 'table']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'stovefan', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'towel', 'wall', 'powersocket', 'photoframe', 'coffee_filter', 'coin', 'tray', 'food_onion', 'doorjamb', 'towel_rack', 'dishrack', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom. Turn on the sink. Rinse hands. Apply soap and scrub your hands together. Rinse hands again. Turn off faucet. Dry hands on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'home_office']\nTask:Check email\nresponse:\n{\n    \"thoughts\": Turn on the computer.  Wait for it to load.  Get online.  Go to the email service.  Open the email.  Read the email.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mail>', '[GRAB] <mail>', '[READ] <mail>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'dog', 'cpuscreen', 'keyboard', 'ceilinglamp', 'plate', 'laptop', 'faucet', 'window', 'stereo', 'oil', 'food_noodles', 'mousepad', 'dining_room', 'nightstand', 'food_food', 'check', 'wine', 'curtain', 'after_shave', 'comb', 'dresser', 'napkin', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I reach out arm. I put hand on radio. I turn on radio. I pick a station.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[TOUCH] <stereo>", "[TURNTO] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'dough', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'window', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'check', 'curtain', 'mouse', 'pencil', 'dresser', 'light', 'table', 'vase', 'shoe_rack']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Put child in bed, sit down,read book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'trashcan', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'bookshelf', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'homework', 'pillow', 'wall', 'pasta', 'powersocket', 'photoframe', 'oven_mitts', 'doorjamb', 'kitchen_counter', 'comb', 'door', 'bathtub', 'light', 'dishwasher']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to bathroom, I open the door of the bathroom, and sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'walllamp', 'bathroom', 'drawing', 'orchid', 'hanger', 'floor', 'phone', 'character', 'stamp', 'tea_bag', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'iron', 'brush', 'window', 'couch', 'food_orange', 'pillow', 'wall', 'bench', 'tray', 'curtain', 'after_shave', 'sink', 'bathroom_cabinet', 'table', 'home_office', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Sitting down and relaxing on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'bookshelf', 'keyboard', 'brush', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'stovefan', 'shoe-shine_kit', 'pillow', 'wall', 'powersocket', 'photoframe', 'novel', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'bathroom_cabinet', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room. Sit on the couch. Turn on TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'mop', 'nail_polish', 'detergent', 'trashcan', 'food_carrot', 'drawing', 'hanger', 'longboard', 'floor', 'keyboard', 'ceilinglamp', 'plate', 'faucet', 'window', 'couch', 'wall', 'nightstand', 'powersocket', 'bench', 'novel', 'food_food', 'television', 'wine', 'doorjamb', 'conditioner', 'bathroom_cabinet', 'table', 'clothes_skirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the sofa. Sit down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_pizza', 'tablelamp', 'mop', 'chair', 'freezer', 'walllamp', 'dough', 'drawing', 'knifeblock', 'wallshelf', 'orchid', 'floor', 'ceiling', 'mop_bucket', 'toaster', 'stovefan', 'faucet', 'bowl', 'wall', 'oven', 'bench', 'toy', 'tray', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "go to the kitchen and cook some thing to eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <oven>", "[OPEN] <oven>", "[FIND] <food_pizza>", "[GRAB] <food_pizza>", "[PUTIN] <food_pizza> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'beer', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'food_orange', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Make toast\nresponse:\n{\n    \"thoughts\": Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[CLOSE] <freezer>', '[FIND] <toaster>', '[PLUGIN] <toaster>', '[PUTBACK] <food_bread> <toaster>', '[SWITCHON] <toaster>']\n}\n", "input": "\nScene_Objects:['walllamp', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'oven', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'home_office', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the Living Room.  I find the remote.  I grab the remote and take a seat on the couch.  I switch on the TV with the remote and  watch the television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'bathroom', 'detergent', 'food_carrot', 'cpuscreen', 'wallshelf', 'phone', 'bedroom', 'desk', 'remote_control', 'envelope', 'computer', 'mail', 'tvstand', 'shower', 'iron', 'mousepad', 'closetdrawer', 'oven', 'powersocket', 'newspaper', 'novel', 'food_food', 'curtain', 'dresser', 'food_steak', 'sink', 'table', 'shoe_rack']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[PUSH] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'jelly', 'bathroom', 'detergent', 'food_butter', 'drawing', 'needle', 'cpuscreen', 'bathroom_counter', 'character', 'stovefan', 'plate', 'window', 'bowl', 'wall', 'bench', 'cup', 'bed', 'coffe_maker', 'check', 'tray', 'towel_rack', 'kitchen_counter', 'vacuum_cleaner', 'pajamas', 'sink', 'light', 'dishwasher', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room. Turn on radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'chair', 'blow_dryer', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'toaster', 'laptop', 'deck_of_cards', 'faucet', 'food_noodles', 'closetdrawer', 'food_orange', 'wall', 'nightstand', 'clothes_gloves', 'coffe_maker', 'toilet_paper', 'food_food', 'curtain', 'towel_rack', 'console', 'kitchen_counter', 'shoes', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Making coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'toilet', 'walllamp', 'detergent', 'food_carrot', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'bedroom', 'keyboard', 'bookshelf', 'mat', 'mousepad', 'dining_room', 'wall', 'nightstand', 'bench', 'basket_for_clothes', 'bed', 'photoframe', 'coffe_maker', 'food_food', 'wooden_spoon', 'mouse', 'dresser', 'sink', 'dishwasher']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "go t the bedroom and go to sleep", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'beer', 'crayon', 'freezer', 'walllamp', 'bathroom', 'detergent', 'wallshelf', 'floor', 'ceiling', 'purse', 'computer', 'mat', 'tvstand', 'shower', 'window', 'couch', 'food_peanut_butter', 'dining_room', 'wall', 'photoframe', 'bed', 'television', 'curtain', 'diary', 'cupboard', 'controller', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Wipe down baseboards please\nresponse:\n{\n    \"thoughts\": Use a rag to get all the dust off the baseboards please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <wall>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'door', 'home_office', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the remote. Pick up the remote. Find the sofa. Sit on the sofa. Look at the television. Aim the remote at the television. Press the power button. Flip the channels.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[PUSH] <remote_control>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'walllamp', 'knife', 'knifeblock', 'cpuscreen', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'tea_bag', 'toaster', 'stovefan', 'plate', 'faucet', 'deck_of_cards', 'tvstand', 'bowl', 'wall', 'oven', 'cup', 'coffe_maker', 'drinking_glass', 'food_donut', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Unload dishwasher\nresponse:\n{\n    \"thoughts\": Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[PUTBACK] <drinking_glass> <cupboard>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <fork> <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTIN] <knife> <kitchen_cabinet>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTIN] <spoon> <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'bowl', 'wall', 'powersocket', 'photoframe', 'food_food', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'shaving_cream', 'towel_rack', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I enter the living room and sit on the couch. I grab the remote and turn on the TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'mop', 'microwave', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'newspaper', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'food_carrot', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'toaster', 'window', 'foundation', 'wall', 'bed', 'food_food', 'coffe_maker', 'toothbrush_holder', 'water', 'drinking_glass', 'food_onion', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'hair', 'door', 'bathroom_cabinet', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'woman', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'man', 'bookshelf', 'keyboard', 'piano_bench', 'stovefan', 'tvstand', 'shower', 'faucet', 'oil', 'dining_room', 'food_kiwi', 'nightstand', 'bench', 'oven', 'powersocket', 'coffe_maker', 'tray', 'curtain', 'towel_rack', 'cupboard', 'door', 'sink', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Stand next to the door. When the door opens from the outside, greet the guests.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <woman>', '[FIND] <woman>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>', '[FIND] <man>', '[TURNTO] <man>', '[LOOKAT] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_bread', 'chair', 'freezer', 'walllamp', 'detergent', 'trashcan', 'drawing', 'desk', 'keyboard', 'bookshelf', 'bookmark', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'pillow', 'bowl', 'wall', 'box', 'coffe_maker', 'razor', 'television', 'curtain', 'doorjamb', 'towel_rack', 'door', 'sink', 'bathroom_cabinet']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Pour water into a glass and drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'chair', 'freezer', 'ground_coffee', 'floor', 'ceiling', 'mail', 'food_bacon', 'toaster', 'deck_of_cards', 'faucet', 'window', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toy', 'light_bulb', 'coffe_maker', 'coffee_filter', 'water', 'curtain', 'after_shave', 'pencil', 'kitchen_counter', 'pajamas', 'sink', 'light']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[FIND] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'microwave', 'ground_coffee', 'knifeblock', 'wallshelf', 'ceiling', 'phone', 'character', 'desk', 'piano_bench', 'stovefan', 'toaster', 'couch', 'clothes_hat', 'pillow', 'bench', 'oven', 'tape', 'cutting_board', 'doorjamb', 'cards', 'mouse', 'dishrack', 'cupboard', 'hairdryer', 'sink', 'clothes_scarf', 'table', 'dishwasher']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "i go find the phone. i pick up the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'band-aids', 'drawing', 'orchid', 'kitchen_cabinet', 'floor', 'bathroom_counter', 'phone', 'bookshelf', 'computer', 'stovefan', 'mat', 'plate', 'ceilinglamp', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'bench', 'powersocket', 'oven', 'cutting_board', 'curtain', 'mouse', 'table', 'dishwasher', 'light']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the kitchen cabinet. Grab the plate and put it on the table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'drawing', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'phone', 'stamp', 'bedroom', 'remote_control', 'button', 'mat', 'faucet', 'homework', 'couch', 'pillow', 'wall', 'basket_for_clothes', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'table', 'light', 'laser_pointer']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "go to living room. grab remote control. sit on couch. point remote at tv. push buttons.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'bathroom', 'trashcan', 'drawing', 'orchid', 'floor', 'phone', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'faucet', 'picture', 'window', 'closetdrawer', 'couch', 'dining_room', 'bed', 'light_bulb', 'keys', 'check', 'video_game_controller', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Petting the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'cup', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "You walk to the living room and sit on the sofa. You grab the remote control, look at it and point at the television, you watch television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'floor', 'character', 'keyboard', 'bookshelf', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'wall_clock', 'toilet_paper', 'mouthwash', 'television', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, find sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'plate', 'faucet', 'tvstand', 'clothes_hat', 'pillow', 'wall', 'oven', 'powersocket', 'newspaper', 'photoframe', 'check', 'coffe_maker', 'doorjamb', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'clothes_scarf', 'table', 'clothes_skirt', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[TURNTO] <plate>", "[LOOKAT] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[WASH] <plate>", "[PUTBACK] <plate> <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'glue', 'toilet', 'chair', 'detergent', 'dough', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'keyboard', 'alcohol', 'stovefan', 'ceilinglamp', 'couch', 'microphone', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'doorjamb', 'kitchen_counter', 'hairbrush', 'cupboard', 'home_office', 'dishwasher', 'coffee', 'shoe_rack']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['juice', 'blender', 'tea', 'toilet', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'keyboard', 'bookshelf', 'toaster', 'ceilinglamp', 'shower', 'mousepad', 'food_jam', 'pillow', 'wall', 'powersocket', 'bed', 'tray', 'kitchen_counter', 'dresser', 'cupboard', 'home_office', 'table', 'bathroom_cabinet']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen and put groceries in fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'freezer', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'brush', 'tvstand', 'stereo', 'window', 'closetdrawer', 'towel', 'wall', 'powersocket', 'tape', 'food_food', 'television', 'curtain', 'doorjamb', 'vacuum_cleaner', 'dresser', 'door', 'clothes_scarf', 'dishwasher', 'board_game']\nTask:vacuum carpet\nresponse:\n{\n    \"thoughts\": I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'jelly', 'drawing', 'orchid', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'computer', 'bookmark', 'faucet', 'clothes_socks', 'couch', 'pillow', 'broom', 'wall', 'wall_clock', 'tape', 'photoframe', 'novel', 'check', 'food_onion', 'doorjamb', 'door', 'bathtub', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to the living room and relax on the sofa with a book", "actions": ["[RUN] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[DROP] <bookmark>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'jelly', 'ceilingfan', 'drawing', 'hanger', 'floor', 'stamp', 'desk', 'ceilinglamp', 'balanceball', 'faucet', 'window', 'couch', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'bed', 'photoframe', 'television', 'oven_mitts', 'food_rice', 'doorjamb', 'cards', 'shaving_cream', 'sink', 'table', 'light', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to living room and lie down on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'bathroom', 'drawing', 'wallshelf', 'cpuscreen', 'character', 'desk', 'envelope', 'computer', 'shower', 'faucet', 'electric_shaver', 'dustpan', 'dining_room', 'wall', 'bench', 'tape', 'photoframe', 'check', 'doorjamb', 'centerpiece', 'mouse', 'hairbrush', 'bathroom_cabinet', 'table', 'home_office', 'light', 'instrument_guitar', 'creditcard']\nTask:Receive credit card\nresponse:\n", "output": {"thoughts": "hear the dooble, go to door, take de envelope, open it with a scissor, take the creditcard", "actions": ["[WALK] <home_office>", "[WALK] <envelope>", "[FIND] <envelope>", "[GRAB] <envelope>", "[WALK] <home_office>", "[WALK] <scissors>", "[FIND] <scissors>", "[OPEN] <envelope>", "[FIND] <creditcard>", "[GRAB] <creditcard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'brush', 'mat', 'tvstand', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'photoframe', 'food_food', 'television', 'doorjamb', 'bag', 'light', 'dishwasher', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'blender', 'band-aids', 'trashcan', 'drawing', 'knifeblock', 'bathroom_counter', 'shampoo', 'phone', 'remote_control', 'mat', 'plate', 'window', 'oil', 'food_jam', 'bowl', 'wall', 'nightstand', 'newspaper', 'keys', 'check', 'coffe_maker', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "take remote control, push \"change channel\" button", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'walllamp', 'ceilingfan', 'food_carrot', 'toothbrush', 'drawing', 'desk', 'bookshelf', 'stovefan', 'balanceball', 'shower', 'window', 'candle', 'clothes_pants', 'food_orange', 'dining_room', 'bench', 'bed', 'keys', 'food_food', 'diary', 'mouse', 'shoes', 'sink', 'vase', 'tooth_paste', 'scrabble', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": EARLY MORNING AND BEFORE GOING TO BED,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'stamp', 'desk', 'button', 'toaster', 'mat', 'plate', 'shower', 'dish_soap', 'pillow', 'wall', 'nightstand', 'food_cheese', 'coffe_maker', 'food_food', 'kitchen_counter', 'dishrack', 'cupboard', 'controller', 'hairdryer', 'home_office', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find dishwasher. Open dishwasher. Find dishrack. Find plate 1. Grab plate 1. Put plate 1 in dishrack. Repeat for dish 2. Find dish soap. Grab dish soap. Put dish soap to dishwasher. Close dishwasher. Switch on dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'blender', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'orchid', 'food_egg', 'kitchen_cabinet', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'wall', 'oven', 'food_cheese', 'food_food', 'food_donut', 'food_rice', 'wine', 'tray', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Walk into kitchen. Open fridge. Take food from fridge. Close fridge. Put food on kitchen counter. Prepare food. Turn on stove. Open kitchen cabinet. Take pot. Close cabinet. Put food in pot. Put pot on stove. Cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <kitchen_counter>', '[PUTBACK] <food_food> <kitchen_counter>', '[CLOSE] <freezer>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[CLOSE] <kitchen_cabinet>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <oven>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'tablelamp', 'freezer', 'bathroom', 'bathroom_counter', 'phone', 'table_cloth', 'toaster', 'mat', 'tvstand', 'faucet', 'mousepad', 'towel', 'nightstand', 'oven', 'cup', 'bed', 'photoframe', 'newspaper', 'food_food', 'after_shave', 'doorjamb', 'towel_rack', 'mouse', 'dishrack', 'controller', 'dishwasher', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into my bedroom, go to the lamp and use the switch to turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[LOOKAT] <floor_lamp>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'dough', 'bills', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'shampoo', 'bookshelf', 'computer', 'mat', 'window', 'food_noodles', 'bowl', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'cup', 'bed', 'toilet_paper', 'mouthwash', 'ice', 'doorjamb', 'vacuum_cleaner', 'dishrack', 'door', 'napkin', 'home_office']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Walk to bedroom. Walk to bed, lie in bed.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'bookshelf', 'mat', 'shower', 'faucet', 'wall', 'powersocket', 'oven', 'light_bulb', 'novel', 'check', 'television', 'food_donut', 'curtain', 'doorjamb', 'kitchen_counter', 'spoon', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Switch on light in home office", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'microwave', 'chair', 'wallshelf', 'hanger', 'floor', 'phone', 'food_bacon', 'piano_bench', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'powersocket', 'television', 'coin', 'food_donut', 'ice', 'rag', 'pencil', 'shoes', 'dresser', 'clothes_dress', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Hang up jacket\nresponse:\n{\n    \"thoughts\": I put clothes on hangers and hang them in the closet,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[FIND] <hanger>', '[OPEN] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'microwave', 'walllamp', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'bedroom', 'ceilinglamp', 'plate', 'faucet', 'window', 'dish_soap', 'mousepad', 'bowl', 'wall', 'oven', 'powersocket', 'check', 'mouthwash', 'after_shave', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I go to the kitchen and turn on the faucet. I get dish soap and put it into the sink. I grab a sponge and a bowl and start cleaning it. I put the bowl into the dishrack. I find a plate and clean it as well. I put it into the dishrack, turn off the faucet and put the sponge back.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <bowl>", "[GRAB] <bowl>", "[POUR] <dish_soap> <sink>", "[PUTOBJBACK] <dish_soap>", "[FIND] <sponge>", "[GRAB] <sponge>", "[FIND] <bowl>", "[SCRUB] <bowl>", "[RINSE] <bowl>", "[FIND] <dishrack>", "[PUTBACK] <bowl> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[SWITCHOFF] <faucet>", "[PUTOBJBACK] <sponge>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'mat', 'stereo', 'faucet', 'mousepad', 'wall', 'wall_clock', 'cup', 'bed', 'photoframe', 'novel', 'coffe_maker', 'television', 'mouse', 'kitchen_counter', 'headset', 'cupboard', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my bedroom.  I locate my stereo and turn it on.  I put on my headphones and sit to listen.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[WALK] <stereo>', '[SWITCHON] <stereo>', '[FIND] <headset>', '[GRAB] <headset>', '[PUTON] <headset>', '[FIND] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'cat', 'toothbrush', 'orchid', 'knife', 'bills', 'bathroom_counter', 'floor', 'desk', 'stovefan', 'laptop', 'shoe-shine_kit', 'shower', 'window', 'mousepad', 'couch', 'powersocket', 'photoframe', 'wooden_spoon', 'curtain', 'doorjamb', 'coffee_pot', 'mouse', 'door', 'bathtub', 'home_office', 'table', 'light', 'clothes_shirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my cat, and I pet him", "actions": ["[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_oatmeal', 'microwave', 'freezer', 'bathroom', 'nail_polish', 'dough', 'food_carrot', 'orchid', 'drawing', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'brush', 'mat', 'couch', 'pillow', 'food_cereal', 'wall', 'food_vegetable', 'food_food', 'television', 'console', 'cards', 'kitchen_counter', 'cupboard', 'hair', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_vegetable> <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTIN] <food_carrot> <freezer>', '[FIND] <food_cereal>', '[GRAB] <food_cereal>', '[PUTIN] <food_cereal> <freezer>', '[FIND] <food_oatmeal>', '[GRAB] <food_oatmeal>', '[PUTIN] <food_oatmeal> <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[FIND] <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'coffee_table', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit down, turn on the TV with the remote control that is on the coffee table.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <coffee_table>", "[TURNTO] <coffee_table>", "[LOOKAT] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'toilet', 'jelly', 'freezer', 'ground_coffee', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'button', 'mat', 'shower', 'picture', 'dustpan', 'food_orange', 'form', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'light_bulb', 'food_food', 'doorjamb', 'cd', 'door', 'light']\nTask:Added meat to freezer\nresponse:\n{\n    \"thoughts\": After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'walllamp', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'folder', 'bedroom', 'food_cake', 'computer', 'shower', 'iron', 'window', 'pillow', 'wall', 'nightstand', 'powersocket', 'tape', 'bed', 'television', 'wine', 'doorjamb', 'towel_rack', 'dresser', 'home_office', 'light', 'bathroom_cabinet', 'dishwasher']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the phone. I pick up the phone. I hold it to my ear.", "actions": ["[WALK] <dining_room>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'food_salt', 'chair', 'walllamp', 'detergent', 'drawing', 'clothes_underwear', 'floor', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'food_noodles', 'garbage_can', 'mousepad', 'microphone', 'wall', 'nightstand', 'food_vegetable', 'bed', 'food_food', 'coffee_filter', 'mouse', 'sink', 'home_office', 'table', 'dishwasher', 'pantry']\nTask:Organize pantry\nresponse:\n{\n    \"thoughts\": Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <pantry>', '[FIND] <food_food>', '[GRAB] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[FIND] <garbage_can>', '[PUTBACK] <food_food> <garbage_can>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <dry_pasta>', '[GRAB] <dry_pasta>', '[TURNTO] <dry_pasta>', '[LOOKAT] <dry_pasta>', '[PUTOBJBACK] <dry_pasta>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[PUSH] <food_food>', '[FIND] <food_noodles>', '[GRAB] <food_noodles>', '[TURNTO] <food_noodles>', '[LOOKAT] <food_noodles>', '[PUTOBJBACK] <food_noodles>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'mop', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'clothes_jacket', 'stovefan', 'faucet', 'window', 'pillow', 'bowl', 'wall', 'wall_clock', 'oven', 'cup', 'photoframe', 'food_food', 'toilet_paper', 'tray', 'after_shave', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'table', 'tooth_paste']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I switch on the stove and cook my groceries there.", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[SWITCHON] <oven>", "[FIND] <bowl>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <oven>", "[FIND] <food_food>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'fork', 'filing_cabinet', 'tablelamp', 'blender', 'mop', 'bathroom', 'detergent', 'toothbrush', 'bills', 'hanger', 'floor', 'bedroom', 'envelope', 'toaster', 'stovefan', 'clothes_socks', 'mousepad', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'toilet_paper', 'towel_rack', 'clothes_dress', 'bag', 'home_office', 'light', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'food_ice_cream', 'microwave', 'toilet', 'walllamp', 'knifeblock', 'drawing', 'floor', 'ceiling', 'desk', 'computer', 'toaster', 'faucet', 'mousepad', 'oil', 'candle', 'bowl', 'wall', 'bed', 'photoframe', 'coffe_maker', 'lighter', 'cards', 'kitchen_counter', 'cupboard', 'door', 'sink', 'dishwasher', 'colander']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen. Take the dishes from the sink. Apply dish washer, scrub and then wash with water.", "actions": ["[WALK] <sink>", "[FIND] <bowl>", "[GRAB] <bowl>", "[FIND] <dishwasher>", "[SCRUB] <dishwasher>", "[RINSE] <bowl>", "[OPEN] <dishwasher>", "[WASH] <bowl>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'check', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'sink', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to the office and sit in a chair, I turn on the computer and grab the mouse. I type on the keyboard and starting working on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'toilet', 'walllamp', 'nail_polish', 'dough', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'cards', 'dishrack', 'door', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "first i will go to living room and then take the remote and swtitch on television after that il change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[TOUCH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'microwave', 'chair', 'bathroom', 'food_butter', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shower', 'foundation', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'box', 'newspaper', 'coffe_maker', 'food_onion', 'doorjamb', 'mouse', 'controller', 'door', 'bathtub', 'table', 'light', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, find the sofa and sit in the sofa and find pillow. lie down in sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <pillow>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'food_butter', 'food_carrot', 'drawing', 'orchid', 'knifeblock', 'knife', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'toaster', 'mat', 'food_noodles', 'couch', 'pillow', 'wall', 'bed', 'photoframe', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'cupboard', 'door', 'bathroom_cabinet', 'table', 'light', 'dirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Relax on the sofa in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'microwave', 'freezer', 'walllamp', 'bathroom', 'orchid', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'mat', 'tvstand', 'faucet', 'window', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'food_cheese', 'bed', 'novel', 'curtain', 'diary', 'dresser', 'bathroom_cabinet']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Put child in bed, sit down,read book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'bathroom', 'trashcan', 'knifeblock', 'orchid', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'phone', 'desk', 'food_cake', 'computer', 'table_cloth', 'tvstand', 'faucet', 'couch', 'oven', 'powersocket', 'cup', 'photoframe', 'check', 'tray', 'curtain', 'wooden_spoon', 'doorjamb', 'door', 'bathtub', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the home office. Pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'freezer', 'walllamp', 'food_carrot', 'desk', 'food_cake', 'remote_control', 'button', 'ceilinglamp', 'faucet', 'stereo', 'window', 'mousepad', 'couch', 'wall_clock', 'bench', 'box', 'cup', 'bed', 'keys', 'newspaper', 'food_food', 'ice', 'after_shave', 'maindoor', 'bathroom_cabinet', 'light', 'vase', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[PUSH] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'chair', 'walllamp', 'knifeblock', 'floor', 'phone', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'purse', 'toaster', 'mat', 'stereo', 'faucet', 'mousepad', 'pillow', 'mouthwash', 'curtain', 'cards', 'shoes', 'cd', 'dishrack', 'dresser', 'door', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Read a book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'homework', 'window', 'wall', 'powersocket', 'oven', 'toy', 'bed', 'toilet_paper', 'television', 'toothbrush_holder', 'food_donut', 'tray', 'doorjamb', 'door', 'sink', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Lift the remote and change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'food_ice_cream', 'microwave', 'freezer', 'walllamp', 'knifeblock', 'floor', 'bookshelf', 'clothes_jacket', 'plate', 'faucet', 'tvstand', 'towel', 'wall', 'nightstand', 'bench', 'coffe_maker', 'television', 'wooden_spoon', 'after_shave', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'pajamas', 'door', 'sink', 'home_office', 'tooth_paste']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "go into the kitchen, walk to the sink. Turn on the water and rinse of a plate and clean them off fully with a sponge.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <plate>", "[RINSE] <plate>", "[FIND] <sponge>", "[GRAB] <sponge>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'measuring_cup', 'jelly', 'freezer', 'nail_polish', 'ceilingfan', 'drawing', 'orchid', 'longboard', 'floor', 'ceiling', 'bedroom', 'keyboard', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'deck_of_cards', 'mousepad', 'couch', 'clothes_pants', 'wall', 'television', 'curtain', 'kitchen_counter', 'chessboard', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk in living room. I find TV remote. I grab TV remote. I sit on sofa. I switch on TV. I switch channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'dry_pasta', 'food_ice_cream', 'toilet', 'freezer', 'knifeblock', 'wallshelf', 'folder', 'desk', 'stamp', 'bedroom', 'computer', 'toaster', 'mat', 'floor_lamp', 'dustpan', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'check', 'curtain', 'console', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go home and it's dark so I can not see anything", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'detergent', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'bathroom_counter', 'phone', 'folder', 'man', 'keyboard', 'mop_bucket', 'tvstand', 'faucet', 'picture', 'plate', 'mousepad', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'check', 'food_onion', 'clothes_scarf', 'spoon', 'light', 'dishwasher']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": When someone knocks I n the door walk through the hallway and greet them.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['sponge', 'toilet', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'ceilinglamp', 'tvstand', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'clothes_gloves', 'novel', 'tray', 'doorjamb', 'kitchen_counter', 'door', 'table', 'light', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'detergent', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'bathroom_counter', 'phone', 'folder', 'man', 'keyboard', 'mop_bucket', 'tvstand', 'faucet', 'picture', 'plate', 'mousepad', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'check', 'food_onion', 'clothes_scarf', 'spoon', 'light', 'dishwasher']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": When someone knocks I n the door walk through the hallway and greet them.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['freezer', 'toothbrush', 'cpuscreen', 'clothes_underwear', 'wallshelf', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'computer', 'button', 'pillow', 'dining_room', 'bench', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'dresser', 'cupboard', 'pajamas', 'clothes_dress', 'sink', 'home_office', 'laser_pointer']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to home office. Walk to computer. Switch on Computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bedroom', 'bookshelf', 'keyboard', 'ceilinglamp', 'faucet', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'food_food', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathtub', 'light', 'table', 'instrument_violin']\nTask:Straighten pictures on wall\nresponse:\n{\n    \"thoughts\": Move the pictures/paintings until they are straight.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PUSH] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PUSH] <drawing>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'ground_coffee', 'walllamp', 'orchid', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'toaster', 'mat', 'faucet', 'wall', 'basket_for_clothes', 'photoframe', 'bed', 'coffe_maker', 'food_food', 'coffee_filter', 'kitchen_counter', 'cd', 'dresser', 'cupboard', 'door', 'napkin', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "You walk into the kitchen and turn to the coffee maker. You grab the coffee filter and put into the coffee machine, you also grab the ground coffee and put it into the coffee machine. You turn on the coffee machine.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'desk', 'bedroom', 'stovefan', 'ceilinglamp', 'plate', 'mat', 'couch', 'wall', 'oven', 'food_food', 'tray', 'food_onion', 'curtain', 'diary', 'kitchen_counter', 'dishrack', 'sauce_pan', 'cupboard', 'spoon', 'home_office']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <spoon>', '[GRAB] <spoon>', '[FIND] <sauce_pan>', '[PUTBACK] <spoon> <sauce_pan>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <plate>', '[PUTBACK] <pot> <plate>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'shower', 'oil', 'couch', 'pillow', 'bowl', 'wall', 'powersocket', 'bed', 'television', 'doorjamb', 'towel_rack', 'kitchen_counter', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'clothes_skirt', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I want to relax on the sofa to some relaxing music in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <stereo>", "[PLUGIN] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'crayon', 'toilet', 'chair', 'freezer', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'clothes_jacket', 'ceilinglamp', 'window', 'pillow', 'wall', 'pasta', 'powersocket', 'toy', 'box', 'cup', 'check', 'coffe_maker', 'toothbrush_holder', 'doorjamb', 'dresser', 'door', 'light', 'scrabble']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTBACK] <toy> <dresser>', '[CLOSE] <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'bathroom', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'toaster', 'stovefan', 'ceilinglamp', 'picture', 'tooth_paste', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'check', 'television', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'home_office', 'bathroom_cabinet', 'board_game']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'chair', 'bathroom', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'character', 'computer', 'mail', 'table_cloth', 'toaster', 'foundation', 'window', 'mousepad', 'bench', 'shelf', 'photoframe', 'bed', 'coffe_maker', 'cards', 'chessboard', 'controller', 'door', 'maindoor', 'bathtub', 'light', 'vase', 'laser_pointer']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk into living room, look at the phone when it ringing, grab the phone or pickup the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'walllamp', 'orchid', 'floor', 'ceiling', 'character', 'desk', 'table_cloth', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'stovefan', 'dish_soap', 'mousepad', 'couch', 'wall', 'powersocket', 'bed', 'coffe_maker', 'food_food', 'kitchen_counter', 'cupboard', 'controller', 'sink', 'dishwasher', 'soap', 'clothes_shirt', 'teeth']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_pizza', 'juice', 'toilet', 'freezer', 'blow_dryer', 'walllamp', 'ceilingfan', 'knifeblock', 'bathroom_counter', 'alcohol', 'button', 'stereo', 'form', 'towel', 'oven', 'basket_for_clothes', 'photoframe', 'light_bulb', 'toilet_paper', 'toothbrush_holder', 'kitchen_counter', 'shoes', 'chessboard', 'vacuum_cleaner', 'clothes_scarf', 'home_office', 'dishwasher', 'soap', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and get the soap and wash my both hand using the soap and rinse my hands and use towel.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[FIND] <towel>', '[RINSE] <towel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'orchid', 'bedroom', 'clothes_jacket', 'mop_bucket', 'stovefan', 'mousepad', 'pillow', 'dining_room', 'nightstand', 'wall', 'powersocket', 'oven', 'basket_for_clothes', 'bed', 'photoframe', 'mouthwash', 'television', 'wine', 'after_shave', 'doorjamb', 'sink', 'bathroom_cabinet', 'light', 'dishwasher', 'soap', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and find the basket of clothes. Put the soap and clothes into the washing machine. Turn on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <washing_machine>", "[TURNTO] <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'brush', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'bench', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'light']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Find the power button. Press.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'food_ice_cream', 'knifeblock', 'needle', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'mop_bucket', 'toaster', 'laptop', 'faucet', 'food_jam', 'wall', 'powersocket', 'box', 'television', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'chessboard', 'door', 'bag', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "approach the light switch, raise hand to the switch, flip the switch", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'mop', 'toilet', 'blow_dryer', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'toaster', 'tvstand', 'shower', 'mousepad', 'pillow', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'bed', 'food_onion', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'bathroom_cabinet']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to bathroom, I open the door of the bathroom, and sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'folder', 'character', 'desk', 'faucet', 'stereo', 'mousepad', 'food_orange', 'wall', 'oven', 'photoframe', 'oven_mitts', 'water', 'drinking_glass', 'wooden_spoon', 'towel_rack', 'shaving_cream', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'spoon', 'board_game']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to kitchen, open cabinet, get glass, close cabinet, put glass under faucet, turn on water, turn off water when glass has 85% water, lift glass to mouth, open mouth and put on glass, tilt glass up and swallow water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[TURNTO] <kitchen_cabinet>", "[LOOKAT] <kitchen_cabinet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[FIND] <sink>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[SWITCHOFF] <faucet>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'measuring_cup', 'band-aids', 'knifeblock', 'cpuscreen', 'dvd_player', 'folder', 'desk', 'bookmark', 'stereo', 'faucet', 'dustpan', 'microphone', 'oven', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'television', 'food_rice', 'pencil', 'shredder', 'controller', 'home_office', 'light', 'table', 'scrabble', 'electrical_outlet']\nTask:Shred receipts\nresponse:\n{\n    \"thoughts\": Take old receipts and put through the shredder,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[FIND] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <check>', '[GRAB] <check>', '[CLOSE] <filing_cabinet>', '[FIND] <electrical_outlet>', '[FIND] <shredder>', '[PLUGIN] <shredder>', '[SWITCHON] <shredder>', '[PUTBACK] <check> <shredder>']\n}\n", "input": "\nScene_Objects:['pot', 'ground_coffee', 'walllamp', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'food_cake', 'toaster', 'window', 'couch', 'wall', 'cup', 'coffe_maker', 'food_food', 'coffee_filter', 'water', 'after_shave', 'towel_rack', 'mouse', 'shaving_cream', 'kitchen_counter', 'cupboard', 'bathtub', 'napkin', 'home_office', 'instrument_guitar', 'dishwasher', 'instrument_violin']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find coffee maker. Find coffee filter. Put it in the coffee maker. Find ground coffee. Put it in the coffee maker. Find Water. Put it in the coffe maker. Switch on the coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <water>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'food_bread', 'filing_cabinet', 'bathroom', 'ceiling', 'phone', 'keyboard', 'purse', 'bookshelf', 'laptop', 'ceilinglamp', 'faucet', 'nightstand', 'bench', 'photoframe', 'bed', 'clothes_gloves', 'food_food', 'coffe_maker', 'television', 'toothbrush_holder', 'curtain', 'shaving_cream', 'pencil', 'dresser', 'conditioner', 'door', 'spoon', 'bathroom_cabinet', 'table']\nTask:Shut front door\nresponse:\n{\n    \"thoughts\": Close the storm door. Lock the storm door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'chair', 'blow_dryer', 'orchid', 'cpuscreen', 'floor', 'phone', 'folder', 'desk', 'bookshelf', 'remote_control', 'computer', 'toaster', 'stovefan', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'photoframe', 'novel', 'television', 'coin', 'coffee_filter', 'doorjamb', 'shoes', 'cupboard', 'sink', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[PUSH] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'microwave', 'toilet', 'freezer', 'trashcan', 'milk', 'drawing', 'floor', 'ceiling', 'phone', 'keyboard', 'mat', 'tvstand', 'shower', 'oil', 'garbage_can', 'form', 'food_dessert', 'food_kiwi', 'wall', 'food_cheese', 'food_vegetable', 'doorjamb', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathroom_cabinet']\nTask:Rotate stock in refrigerator\nresponse:\n{\n    \"thoughts\": Throw out expired items, put older items in front.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[TURNTO] <food_bread>', '[LOOKAT] <food_bread>', '[PUTOBJBACK] <food_bread>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[FIND] <garbage_can>', '[PUTBACK] <food_cheese> <garbage_can>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[PUTOBJBACK] <food_cheese>', '[FIND] <food_dessert>', '[TURNTO] <food_dessert>', '[LOOKAT] <food_dessert>', '[PUSH] <food_dessert>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <milk>', '[GRAB] <milk>', '[TURNTO] <milk>', '[LOOKAT] <milk>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['walllamp', 'detergent', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'coffee_filter', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'sink', 'bathtub', 'light', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I pick up remote. I turn on tv. I push on the remote. I change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'phone', 'remote_control', 'food_bacon', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'powersocket', 'oven', 'cup', 'bed', 'coffe_maker', 'television', 'curtain', 'lighter', 'doorjamb', 'shaving_cream', 'headset', 'cupboard', 'table', 'clothes_shirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": We recently got a 43 inch TV and have enjoyed watching some of our favorite movies and TV shows. We use a remote to do so.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'purse', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'deck_of_cards', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'door', 'bathtub', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room. Sit down on the sofa. Pick up the remote control. Turn on the television with the remote control. Put down the remote control. Watch the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'pot', 'juice', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'drawing', 'cpuscreen', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'bookshelf', 'laptop', 'tvstand', 'faucet', 'electric_shaver', 'dining_room', 'wall', 'nightstand', 'oven', 'photoframe', 'novel', 'water', 'kitchen_counter', 'cupboard', 'hairdryer', 'sink', 'door']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Opening the kitchen cabinet to grab water glass and pour water from the faucet into it to drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[CLOSE] <kitchen_cabinet>', '[FIND] <sink>', '[PUTBACK] <water_glass> <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[SWITCHOFF] <faucet>', '[GRAB] <water_glass>', '[FIND] <water>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['trashcan', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'tea_bag', 'keyboard', 'food_bacon', 'stovefan', 'tvstand', 'faucet', 'couch', 'microphone', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'oven_mitts', 'doorjamb', 'mouse', 'door', 'bathtub', 'sink', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'mop', 'freezer', 'blow_dryer', 'bathroom', 'detergent', 'milk', 'bathroom_counter', 'ceiling', 'character', 'mop_bucket', 'stovefan', 'shoe-shine_kit', 'balanceball', 'deck_of_cards', 'stereo', 'shower', 'mousepad', 'microphone', 'pillow', 'light_bulb', 'coffee_filter', 'doorjamb', 'towel_rack', 'pencil', 'cupboard', 'light', 'dishwasher']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'tablelamp', 'tea', 'microwave', 'chair', 'drawing', 'orchid', 'floor', 'computer', 'food_bacon', 'table_cloth', 'stovefan', 'toaster', 'homework', 'mousepad', 'couch', 'powersocket', 'oven', 'bench', 'check', 'oven_mitts', 'coffee_filter', 'curtain', 'rag', 'doorjamb', 'kitchen_counter', 'shoes', 'bathtub', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room and I pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'sponge', 'beer', 'chair', 'freezer', 'detergent', 'food_butter', 'cpuscreen', 'hanger', 'bedroom', 'desk', 'bookshelf', 'table_cloth', 'toaster', 'ceilinglamp', 'shower', 'iron', 'window', 'wall', 'oven', 'food_food', 'razor', 'food_rice', 'curtain', 'kitchen_counter', 'food_steak', 'door', 'spoon', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'microwave', 'chair', 'bathroom', 'food_butter', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'shower', 'foundation', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'box', 'newspaper', 'coffe_maker', 'food_onion', 'doorjamb', 'mouse', 'controller', 'door', 'bathtub', 'table', 'light', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room, find the sofa and sit in the sofa and find pillow. lie down in sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <pillow>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'beer', 'microwave', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'dough', 'orchid', 'dvd_player', 'keyboard', 'piano_bench', 'brush', 'faucet', 'pillow', 'dining_room', 'powersocket', 'newspaper', 'coffe_maker', 'television', 'wine', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'conditioner', 'clothes_scarf', 'bathtub', 'bathroom_cabinet', 'creditcard']\nTask:Lock door\nresponse:\n{\n    \"thoughts\": Turn the lock.,\n    \"actions\": ['[WALK] <bedroom>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'crayon', 'chair', 'walllamp', 'knifeblock', 'bills', 'wallshelf', 'bathroom_counter', 'phone', 'ceiling', 'bedroom', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'dining_room', 'nightstand', 'oven', 'television', 'tray', 'curtain', 'towel_rack', 'door', 'bathtub', 'napkin', 'bag']\nTask:Read book\nresponse:\n", "output": {"thoughts": "walk to living room ,find a book ,open the book ,read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'beer', 'jelly', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'clothes_jacket', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'food_noodles', 'couch', 'wall', 'oven', 'basket_for_clothes', 'cup', 'television', 'curtain', 'conditioner', 'maindoor', 'table', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit on the couch. Turn on TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'chair', 'trashcan', 'drawing', 'clothes_underwear', 'floor', 'phone', 'character', 'keyboard', 'mat', 'ceilinglamp', 'plate', 'tvstand', 'shower', 'faucet', 'mousepad', 'couch', 'wall', 'bench', 'powersocket', 'newspaper', 'bed', 'television', 'mouse', 'controller', 'table', 'light', 'instrument_violin', 'creditcard']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.", "actions": ["[WALK] <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'window', 'mousepad', 'wall', 'nightstand', 'powersocket', 'bench', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'cd', 'sink', 'bathtub', 'table', 'dirt']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I go to my office and sit at my desk. Turn on the computer and type my email at the keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'nail_polish', 'knifeblock', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'desk', 'cleaning_solution', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'shower', 'faucet', 'ceilinglamp', 'pillow', 'bench', 'cup', 'mouse', 'controller', 'laundry_detergent', 'door', 'napkin', 'light', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[SQUEEZE] <cleaning_solution>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'glue', 'toilet', 'chair', 'detergent', 'dough', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'keyboard', 'alcohol', 'stovefan', 'ceilinglamp', 'couch', 'microphone', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'doorjamb', 'kitchen_counter', 'hairbrush', 'cupboard', 'home_office', 'dishwasher', 'coffee', 'shoe_rack']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'crayon', 'microwave', 'dog', 'band-aids', 'walllamp', 'trashcan', 'knifeblock', 'orchid', 'floor', 'desk', 'bookshelf', 'computer', 'piano_bench', 'plate', 'faucet', 'dish_soap', 'food_jam', 'form', 'wall', 'photoframe', 'doorjamb', 'kettle', 'kitchen_counter', 'dishrack', 'cupboard', 'pajamas', 'sink', 'table']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and turn to the sink. Put the plate into the sink. Grab and squeeze the dish soap. Scrub the plate then rinse it. Put the plate into the dish rack.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <sink>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[SQUEEZE] <dish_soap>", "[SCRUB] <plate>", "[RINSE] <plate>", "[GRAB] <plate>", "[FIND] <dishrack>", "[PUTBACK] <plate> <dishrack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'ground_coffee', 'trashcan', 'desk', 'bedroom', 'purse', 'mop_bucket', 'shower', 'picture', 'homework', 'mousepad', 'form', 'wall', 'oven', 'bed', 'keys', 'video_game_controller', 'wine', 'food_onion', 'after_shave', 'console', 'mouse', 'cd', 'comb', 'dresser', 'door', 'clothes_skirt', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find the cat; pet the cat by touching it,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['glue', 'crayon', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'ceiling', 'character', 'bedroom', 'mat', 'faucet', 'window', 'mousepad', 'couch', 'candle', 'pillow', 'bench', 'oven', 'coffe_maker', 'rag', 'towel_rack', 'dresser', 'pajamas', 'cupboard', 'bathtub', 'napkin', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "You walk to the living room and find the cat. You look at the cat and touch it to pet it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'photoframe', 'novel', 'television', 'mouse', 'kitchen_counter', 'door', 'sink', 'clothes_skirt', 'clothes_shirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to home office, find the book. Sit in the chair and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'floor', 'ceiling', 'phone', 'bedroom', 'bookmark', 'mat', 'shoe-shine_kit', 'tvstand', 'window', 'couch', 'pillow', 'towel', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'coffe_maker', 'mouthwash', 'tray', 'doorjamb', 'controller', 'door', 'light', 'creditcard']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I go to the bedroom and turn on the light. I lie in bed. I grab my book from the nightstand, and look for the bookmark to check where I last left off. I read for a while. When I'm done, I put the bookmark on the page I stop at and place the book back on the nightstand.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[READ] <novel>", "[PUTBACK] <bookmark> <novel>", "[FIND] <nightstand>", "[PUTBACK] <novel> <nightstand>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'bedroom', 'mop_bucket', 'faucet', 'balanceball', 'towel', 'wall', 'photoframe', 'keys', 'novel', 'razor', 'television', 'diary', 'kettle', 'coffee_pot', 'cd', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom and locate the sink. I turn the faucet so that water is running, and rinse my hands briefly. Then I grab the soap and squeeze some onto my hands. I spend a few minutes rubbing my hands together, then I rinse them again. Finally, I turn the faucet off and wipe my hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SQUEEZE] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'glue', 'tablelamp', 'microwave', 'toilet', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'phone', 'food_cake', 'table_cloth', 'plate', 'faucet', 'stereo', 'dish_soap', 'towel', 'bowl', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'centerpiece', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'spoon']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "I enter the kitchen, walk to the sink, turn on the faucet, grab the sponge and get it wet, then apply dish soap to the sponge. I take a dish, scrub it with the sponge, rinse it, and place it on the dish rack to dry, and repeat until all the dishes are clean. I rinse the sponge, put it down, and turn off the faucet.", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <sponge>", "[FIND] <dish_soap>", "[GRAB] <sponge>", "[RINSE] <sponge>", "[GRAB] <dish_soap>", "[POUR] <dish_soap> <sponge>", "[PUTOBJBACK] <dish_soap>", "[SQUEEZE] <sponge>", "[FIND] <bowl>", "[GRAB] <bowl>", "[SCRUB] <bowl>", "[RINSE] <bowl>", "[FIND] <dishrack>", "[PUTBACK] <bowl> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <spoon>", "[GRAB] <spoon>", "[SCRUB] <spoon>", "[RINSE] <spoon>", "[PUTBACK] <spoon> <dishrack>", "[RINSE] <sponge>", "[SQUEEZE] <sponge>", "[PUTOBJBACK] <sponge>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'powersocket', 'oven', 'coffe_maker', 'television', 'coffee_filter', 'food_rice', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'dirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I like to set on my sofa.Turn on the tv. And watch Perry Mason. Monday's through Friday's.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'mop', 'microwave', 'toilet', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'computer', 'piano_bench', 'ceilinglamp', 'plate', 'deck_of_cards', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'bed', 'novel', 'television', 'doorjamb', 'door', 'coffee_cup', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'purse', 'toaster', 'faucet', 'shower', 'homework', 'pillow', 'wall', 'oven', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to kitchen, turn on water, find soap, apply to hand, lather, rinse, dry hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'spectacles', 'filing_cabinet', 'mop', 'toilet', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'desk', 'computer', 'toaster', 'ceilinglamp', 'mat', 'couch', 'clothes_pants', 'pillow', 'food_food', 'television', 'doorjamb', 'towel_rack', 'kitchen_counter', 'hairbrush', 'hairdryer', 'bathtub', 'dishwasher', 'dirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the living room, Find Phone stand. Find Phone. Pickup the receiver to attend the call.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'microwave', 'toilet', 'drawing', 'wallshelf', 'ceiling', 'desk', 'bookmark', 'toaster', 'stovefan', 'mat', 'shower', 'couch', 'wall', 'oven', 'coffe_maker', 'coin', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'cupboard', 'controller', 'hairdryer', 'spoon', 'dishwasher', 'creditcard']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen, take a glass from the cupboard, fill with water and drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>', '[RINSE] <water_glass>', '[PUTBACK] <water_glass> <cupboard>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'towel_rack', 'cd', 'door', 'table', 'light', 'clothes_shirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Pickup remote control for the living room TV and push the change channel button", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'detergent', 'trashcan', 'dough', 'clothes_underwear', 'folder', 'stamp', 'bookshelf', 'clothes_jacket', 'table_cloth', 'mousepad', 'couch', 'food_jam', 'clothes_hat', 'pillow', 'nightstand', 'bench', 'oven', 'cutting_board', 'bed', 'food_food', 'check', 'doorjamb', 'console', 'cupboard', 'home_office', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": i go find the phone. i pick up the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['food_bread', 'drawing', 'orchid', 'knifeblock', 'bills', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'shower', 'stereo', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'television', 'food_rice', 'tray', 'curtain', 'doorjamb', 'pencil', 'kitchen_counter', 'door', 'home_office', 'table', 'light', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the sofa. Sit down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'mop', 'toilet', 'freezer', 'walllamp', 'bathroom', 'cat', 'food_butter', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'iron', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'toy', 'tape', 'photoframe', 'television', 'food_rice', 'doorjamb', 'cards', 'door']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'bookshelf', 'piano_bench', 'stovefan', 'ceilinglamp', 'faucet', 'electric_shaver', 'window', 'teeth', 'food_orange', 'pillow', 'wall', 'powersocket', 'food_food', 'tray', 'curtain', 'dresser', 'hair', 'controller', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk in to the bathroom. I find the soap. I find the faucet. I turn on the faucet. I wash my hands. I put the soap back. I rinse my hands. I turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'juice', 'measuring_cup', 'microwave', 'chair', 'freezer', 'band-aids', 'walllamp', 'detergent', 'knifeblock', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'mop_bucket', 'toaster', 'stovefan', 'faucet', 'couch', 'wall', 'oven', 'cup', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'food_snack', 'sink', 'napkin', 'home_office']\nTask:Have snack\nresponse:\n{\n    \"thoughts\": Pour soda into a glass. Get pretzels out of bag. Eat and drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTIN] <juice> <freezer>', '[CLOSE] <freezer>', '[WALK] <cupboard>', '[OPEN] <cupboard>', '[FIND] <food_snack>', '[GRAB] <food_snack>', '[CLOSE] <cupboard>', '[DRINK] <cup>', '[EAT] <food_snack>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'purse', 'toaster', 'faucet', 'shower', 'homework', 'pillow', 'wall', 'oven', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to kitchen, turn on water, find soap, apply to hand, lather, rinse, dry hands", "actions": ["[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[PUTOBJBACK] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'button', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'faucet', 'mousepad', 'dining_room', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'rag', 'doorjamb', 'kitchen_counter', 'light', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'ground_coffee', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'food_ice_cream', 'toothbrush', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'purse', 'faucet', 'window', 'towel', 'wall', 'toy', 'bed', 'light_bulb', 'novel', 'coffe_maker', 'television', 'water', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'dresser', 'maindoor', 'sink', 'tooth_paste', 'clothes_shirt', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[FIND] <toothbrush>', '[FIND] <water>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'freezer', 'walllamp', 'food_carrot', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'toaster', 'shower', 'wall', 'novel', 'food_food', 'television', 'oven_mitts', 'tray', 'kitchen_counter', 'headset', 'dresser', 'cupboard', 'hair', 'bathtub']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen, open fridge, get groceries, put groceries in fridge, close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'food_salt', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'powersocket', 'toy', 'tape', 'light_bulb', 'check', 'food_food', 'television', 'curtain', 'doorjamb', 'mouse', 'pencil', 'dresser', 'conditioner', 'spoon', 'light', 'table', 'colander']\nTask:Re arrange office\nresponse:\n{\n    \"thoughts\": Walk in.Turn on light.Move furniture. Change the room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[PUSH] <desk>', '[WALK] <chair>', '[PUSH] <chair>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['tea', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'form', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'cupboard', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to livingroom, sit on couch, grab remote and turn on tv.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'photoframe', 'novel', 'television', 'mouse', 'kitchen_counter', 'door', 'sink', 'clothes_skirt', 'clothes_shirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to home office, find the book. Sit in the chair and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'water_glass', 'pot', 'walllamp', 'bathroom', 'milk', 'knife', 'knifeblock', 'drawing', 'floor', 'ceiling', 'bedroom', 'toaster', 'stovefan', 'plate', 'shower', 'faucet', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'coffee_cup', 'dishwasher', 'table']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and put plates, cutlery and glasses into the dishwasher, put the soap, close the machine and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['drawing', 'hanger', 'floor', 'ceiling', 'computer', 'brush', 'mat', 'ceilinglamp', 'faucet', 'stovefan', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cutting_board', 'shelf', 'toy', 'standingmirror', 'television', 'curtain', 'doorjamb', 'towel_rack', 'controller', 'home_office', 'table', 'soap', 'shoe_rack']\nTask:Turn off TV\nresponse:\n{\n    \"thoughts\": Walk over to tv and press power button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['band-aids', 'walllamp', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'button', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'clothes_pants', 'dining_room', 'wall', 'powersocket', 'photoframe', 'coffee_table', 'bed', 'clothes_gloves', 'television', 'curtain', 'doorjamb', 'bathtub', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Go to the living room, find the remote control on the coffee table, and change the TV channel.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <coffee_table>", "[WALK] <coffee_table>", "[TURNTO] <coffee_table>", "[LOOKAT] <coffee_table>", "[FIND] <remote_control>", "[TOUCH] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[TOUCH] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'dry_pasta', 'filing_cabinet', 'scissors', 'kitchen_cabinet', 'envelope', 'faucet', 'dish_soap', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall_clock', 'bench', 'toy', 'cutting_board', 'razor', 'food_onion', 'doorjamb', 'towel_rack', 'kitchen_counter', 'pajamas', 'clothes_dress', 'bathtub', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'vase', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTIN] <dish_soap> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'microwave', 'detergent', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'alcohol', 'toaster', 'shower', 'couch', 'clothes_hat', 'pillow', 'wall', 'pasta', 'bench', 'oven', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'doorjamb', 'hairbrush', 'cupboard', 'door', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'wallshelf', 'bookshelf', 'clothes_jacket', 'toaster', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'dining_room', 'nightstand', 'wall', 'wall_clock', 'powersocket', 'basket_for_clothes', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'console', 'dishrack', 'hairdryer', 'door', 'bathtub']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'envelope', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Enter the living room, find TV remote, grab TV remote, switch on TV, find sofa, sit down on the sofa, watch TV", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'bathroom', 'orchid', 'wallshelf', 'ceiling', 'phone', 'stamp', 'bookshelf', 'mail', 'computer', 'tvstand', 'shower', 'couch', 'pillow', 'oven', 'powersocket', 'box', 'bed', 'clothes_gloves', 'photoframe', 'check', 'curtain', 'pencil', 'dresser', 'cupboard', 'door', 'sink', 'bathtub', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find a cat, touch it gently,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'jelly', 'freezer', 'walllamp', 'knifeblock', 'orchid', 'wallshelf', 'bathroom_counter', 'ceiling', 'bookshelf', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'window', 'clothes_hat', 'nightstand', 'wall', 'bench', 'coffe_maker', 'tray', 'curtain', 'doorjamb', 'mouse', 'cupboard', 'napkin', 'bathroom_cabinet', 'clothes_skirt', 'instrument_violin']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drink water", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'chair', 'bathroom', 'detergent', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'mat', 'tvstand', 'clothes_pants', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'novel', 'check', 'keys', 'tray', 'after_shave', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'tooth_paste']\nTask:Read\nresponse:\n{\n    \"thoughts\": I like to read books in my bedroom.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <desk>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <bed>', '[LIE] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'remote_control', 'keyboard', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'faucet', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'light', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I grab the remote.  I find the red power circular button. I point the remote at the television.   I push button down and TV screen lights up. I sit and watch.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'chair', 'bathroom_counter', 'floor', 'laptop', 'faucet', 'shower', 'towel', 'wall', 'wall_clock', 'nightstand', 'cup', 'bed', 'photoframe', 'water', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go into the bathroom and go over to the sink. Grab soap and turn on the water. Wet your hands and rub them together so that the soap lathers. Rub the soap all over your hands thoroughly. Rinse off the soap with water and turn off the sink. Dry your hands with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <hands_both>', '[POUR] <water> <hands_both>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'sink', 'light', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room ,find television , switch on remote control", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[SWITCHON] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'band-aids', 'walllamp', 'detergent', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'video_game_console', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'plate', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'photoframe', 'video_game_controller', 'television', 'curtain', 'dresser', 'table', 'dishwasher']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on game.  Use controller,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[PLUGIN] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'cd_player', 'blow_dryer', 'bathroom', 'detergent', 'orchid', 'bills', 'bathroom_counter', 'phone', 'bedroom', 'remote_control', 'thread', 'button', 'mat', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'broom', 'wall_clock', 'powersocket', 'bench', 'box', 'coffe_maker', 'razor', 'towel_rack', 'vacuum_cleaner', 'bathtub', 'dishwasher', 'soap']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[PLUGIN] <cd_player>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'knifeblock', 'needle', 'bills', 'drawing', 'phone', 'mat', 'deck_of_cards', 'shoe-shine_kit', 'ceilinglamp', 'electric_shaver', 'window', 'clothes_socks', 'couch', 'dining_room', 'nightstand', 'wall', 'bench', 'cup', 'bed', 'mouthwash', 'coffee_filter', 'rag', 'towel_rack', 'hairbrush', 'dresser', 'clothes_dress', 'bathroom_cabinet', 'laser_pointer']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to phone, grab phone, pick up phone, put to ear,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'detergent', 'trashcan', 'drawing', 'food_egg', 'ceiling', 'keyboard', 'stovefan', 'mat', 'ceilinglamp', 'toaster', 'window', 'mousepad', 'couch', 'dining_room', 'oven', 'light_bulb', 'check', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'mouse', 'cupboard', 'door', 'bathtub', 'table', 'dishwasher', 'clothes_skirt', 'creditcard']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the living room, walk over to phone, grab the phone, put it to your ear .", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tea', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'phone', 'keyboard', 'food_cake', 'toaster', 'stovefan', 'plate', 'mat', 'dish_soap', 'oil', 'pillow', 'wall', 'cup', 'photoframe', 'coffe_maker', 'curtain', 'centerpiece', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'chair', 'bathroom', 'detergent', 'trashcan', 'toothbrush', 'knifeblock', 'bathroom_counter', 'character', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'nightstand', 'oven', 'bench', 'powersocket', 'cup', 'bed', 'ice', 'diary', 'cards', 'dresser', 'hairbrush', 'home_office', 'dishwasher', 'scrabble']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "go to radio. run on radio.", "actions": ["[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'toothbrush', 'cpuscreen', 'clothes_underwear', 'wallshelf', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'computer', 'button', 'pillow', 'dining_room', 'bench', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'dresser', 'cupboard', 'pajamas', 'clothes_dress', 'sink', 'home_office', 'laser_pointer']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on Computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'laptop', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watching TV", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'toilet', 'freezer', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'shoe-shine_kit', 'shower', 'faucet', 'mat', 'oil', 'food_peanut_butter', 'towel', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'toilet_paper', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'sink', 'bathroom_cabinet', 'light', 'colander']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <bathroom_cabinet>', '[FIND] <shower>', '[RINSE] <shower>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'microwave', 'freezer', 'drawing', 'phone', 'bookshelf', 'toaster', 'stovefan', 'tvstand', 'mousepad', 'couch', 'clothes_hat', 'food_orange', 'wall', 'photoframe', 'clothes_gloves', 'bed', 'check', 'television', 'food_rice', 'tray', 'doorjamb', 'dresser', 'hairdryer', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'door']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to my bedroom, and then I switch on the radio", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'beer', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'milk', 'drawing', 'floor', 'ceiling', 'toaster', 'mat', 'shower', 'mousepad', 'garbage_can', 'food_dessert', 'wall', 'nightstand', 'powersocket', 'oven', 'food_vegetable', 'food_cheese', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'laundry_detergent', 'door', 'light']\nTask:Rotate stock in refrigerator\nresponse:\n{\n    \"thoughts\": Throw out expired items, put older items in front.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[TURNTO] <food_bread>', '[LOOKAT] <food_bread>', '[PUTOBJBACK] <food_bread>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[FIND] <garbage_can>', '[PUTBACK] <food_cheese> <garbage_can>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[PUTOBJBACK] <food_cheese>', '[FIND] <food_dessert>', '[TURNTO] <food_dessert>', '[LOOKAT] <food_dessert>', '[PUSH] <food_dessert>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <milk>', '[GRAB] <milk>', '[TURNTO] <milk>', '[LOOKAT] <milk>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'keyboard', 'computer', 'stovefan', 'brush', 'window', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'oven', 'coffe_maker', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'shaving_cream', 'cards', 'door', 'table', 'light', 'tooth_paste', 'laser_pointer']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "The cat is laying on the couch in the living room. I want to pet it, so I am going to walk to the living room, sit on the couch, and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_salt', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'mousepad', 'microphone', 'dining_room', 'nightstand', 'basket_for_clothes', 'bed', 'newspaper', 'coffe_maker', 'check', 'tray', 'curtain', 'doorjamb', 'mouse', 'headset', 'laundry_detergent', 'maindoor', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to laundry room. Find laundry basket. Grab laundry basket. Find washing machine. Walk to washing machine. Open washing machine. Find detergent. Grab detergent. Put in detergent. Put in clothes. Close washing machine. Turn on machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[PUTOBJBACK] <laundry_detergent>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'chair', 'detergent', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'shower', 'mousepad', 'dustpan', 'dining_room', 'wall', 'powersocket', 'oven', 'check', 'television', 'doorjamb', 'hairbrush', 'controller', 'sink', 'home_office', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find light switch.  Switch on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'toilet', 'walllamp', 'nail_polish', 'food_butter', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'ceilinglamp', 'tvstand', 'shower', 'closetdrawer', 'couch', 'form', 'dining_room', 'wall', 'wall_clock', 'bed', 'toothbrush_holder', 'doorjamb', 'chessboard', 'door', 'napkin', 'bathroom_cabinet', 'shoe_rack']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk over to the phone and pick it up.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'toilet', 'freezer', 'ground_coffee', 'knifeblock', 'floor', 'ceiling', 'character', 'bedroom', 'keyboard', 'purse', 'computer', 'toaster', 'window', 'mousepad', 'broom', 'wall', 'check', 'food_food', 'food_rice', 'doorjamb', 'mouse', 'pencil', 'dresser', 'sink', 'napkin', 'light', 'colander', 'tooth_paste', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "grab the vesseles and wash", "actions": ["[WALK] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dining_room>", "[WASH] <bowl>", "[WALK] <detergent>", "[FIND] <detergent>", "[RINSE] <detergent>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'bathroom', 'trashcan', 'dough', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'faucet', 'window', 'food_noodles', 'mousepad', 'pillow', 'dining_room', 'wall', 'bed', 'food_food', 'razor', 'rag', 'console', 'mouse', 'towel_rack', 'kitchen_counter', 'door', 'sink', 'bathtub', 'bathroom_cabinet']\nTask:Clean sink\nresponse:\n{\n    \"thoughts\": Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[WIPE] <faucet>', '[WIPE] <sink>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'computer', 'toaster', 'faucet', 'shower', 'closetdrawer', 'couch', 'wall', 'nightstand', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'hairbrush', 'laundry_detergent', 'door']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug in vacuum. Walk around with vacuum, covering all of the floor. Unplug vacuum.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'beer', 'detergent', 'food_butter', 'orchid', 'hanger', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'closetdrawer', 'broom', 'wall', 'nightstand', 'toy', 'newspaper', 'toilet_paper', 'television', 'wine', 'doorjamb', 'kitchen_counter', 'shoes', 'chessboard', 'dresser', 'door', 'dishwasher', 'instrument_violin', 'laser_pointer']\nTask:Put away shoes\nresponse:\n{\n    \"thoughts\": Pick up shoes from entry hall. Open entry way closet. Put shoes in closet. Close closet door.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[FIND] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'food_salt', 'microwave', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'keyboard', 'food_cake', 'purse', 'shoe-shine_kit', 'faucet', 'wall', 'nightstand', 'powersocket', 'photoframe', 'check', 'coffe_maker', 'doorjamb', 'hair', 'sink', 'bathroom_cabinet', 'table']\nTask:Clean sink\nresponse:\n", "output": {"thoughts": "Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.", "actions": ["[WALK] <bathroom>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[WIPE] <faucet>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'chair', 'jelly', 'walllamp', 'drawing', 'wallshelf', 'food_egg', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'food_rice', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'coffee']\nTask:Watch  horror  movie\nresponse:\n{\n    \"thoughts\": I would  find something  scary  to  watch on dvd,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'food_salt', 'dog', 'freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'bedroom', 'computer', 'toaster', 'mat', 'tvstand', 'clothes_socks', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'dining_room', 'wall', 'tape', 'check', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'table']\nTask:Put out flowers\nresponse:\n", "output": {"thoughts": "Place a flower centrepiece on the entrance hall table.", "actions": ["[WALK] <home_office>", "[WALK] <centerpiece>", "[FIND] <centerpiece>", "[GRAB] <centerpiece>", "[WALK] <table>", "[PUTBACK] <centerpiece> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'button', 'piano_bench', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'novel', 'television', 'doorjamb', 'dresser']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'table', 'coffee']\nTask:Call family member with skype application\nresponse:\n", "output": {"thoughts": "Sit down on desk chair located in front of computer.  Turn on computer.  Log in to skype application.  Locate individual to call on contacts list.  Press the call button with computer mouse.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[POINTAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'filing_cabinet', 'tablelamp', 'crayon', 'woman', 'wallshelf', 'shampoo', 'floor', 'phone', 'desk', 'man', 'bookshelf', 'closetdrawer', 'couch', 'dining_room', 'nightstand', 'child', 'bed', 'photoframe', 'standingmirror', 'toilet_paper', 'oven_mitts', 'towel_rack', 'kitchen_counter', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'vase']\nTask:Say goodbye to guests leaving\nresponse:\n{\n    \"thoughts\": Tell them bye,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[FIND] <man>', '[TOUCH] <man>', '[FIND] <woman>', '[TOUCH] <woman>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'drawing', 'orchid', 'wallshelf', 'floor', 'desk', 'bookshelf', 'purse', 'computer', 'mat', 'tvstand', 'couch', 'clothes_pants', 'pillow', 'wall', 'bench', 'powersocket', 'basket_for_clothes', 'photoframe', 'bed', 'television', 'doorjamb', 'pencil', 'laundry_detergent', 'clothes_dress', 'table', 'light', 'clothes_shirt', 'washing_machine']\nTask:Do laundry\nresponse:\n", "output": {"thoughts": "Collecting laundry from the children's hamper", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <basket_for_clothes> <table>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[POUR] <laundry_detergent> <washing_machine>", "[PUTOBJBACK] <laundry_detergent>", "[CLOSE] <filing_cabinet>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[PLUGIN] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'chair', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'stereo', 'floor_lamp', 'mousepad', 'wall', 'nightstand', 'novel', 'video_game_controller', 'television', 'mouse', 'door', 'napkin']\nTask:Study\nresponse:\n{\n    \"thoughts\": Study norwegian,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'freezer', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'mail', 'stovefan', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'pillow', 'wall_clock', 'child', 'check', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'dresser', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Say hi", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'bathroom', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'alcohol', 'mat', 'electric_shaver', 'oil', 'wall', 'powersocket', 'box', 'food_cheese', 'photoframe', 'bed', 'food_food', 'coffe_maker', 'doorjamb', 'lighter', 'dishrack', 'door', 'bathroom_cabinet', 'light', 'dishwasher', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": push any ligth button to turn it on, for example in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['spectacles', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'newspaper', 'television', 'curtain', 'rag', 'doorjamb', 'mouse', 'door', 'bathtub', 'table']\nTask:Use computer\nresponse:\n", "output": {"thoughts": "Browse reddit", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'bookshelf', 'mat', 'shower', 'faucet', 'wall', 'powersocket', 'oven', 'light_bulb', 'novel', 'check', 'television', 'food_donut', 'curtain', 'doorjamb', 'kitchen_counter', 'spoon', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Switch on light in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[SWITCHON] <light>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'toaster', 'stovefan', 'faucet', 'towel', 'pillow', 'wall', 'wall_clock', 'bed', 'check', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'controller', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Cleaning hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'toilet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'mail', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'cutting_board', 'photoframe', 'bed', 'television', 'doorjamb', 'dishrack', 'dresser', 'door', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": go to the living room to watch TV,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[TOUCH] <remote_control>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[DROP] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'freezer', 'knife', 'food_egg', 'wallshelf', 'kitchen_cabinet', 'ceiling', 'character', 'desk', 'tea_bag', 'toaster', 'stovefan', 'plate', 'faucet', 'shower', 'mat', 'wall', 'oven', 'bench', 'coffe_maker', 'drinking_glass', 'tray', 'kitchen_counter', 'cupboard', 'door', 'sink', 'spoon', 'napkin', 'dishwasher']\nTask:Unload dishwasher\nresponse:\n", "output": {"thoughts": "Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[PUTBACK] <drinking_glass> <cupboard>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[PUTIN] <fork> <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTIN] <knife> <kitchen_cabinet>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTIN] <spoon> <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTIN] <plate> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'jelly', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'homework', 'closetdrawer', 'couch', 'pillow', 'wall', 'newspaper', 'coffe_maker', 'television', 'tray', 'curtain', 'towel_rack', 'cupboard', 'home_office', 'table']\nTask:Read newspaper\nresponse:\n{\n    \"thoughts\": Read the newspaper on the couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <newspaper>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'walllamp', 'bathroom', 'detergent', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'stovefan', 'mat', 'tvstand', 'faucet', 'shower', 'ceilinglamp', 'couch', 'towel', 'dining_room', 'wall', 'oven', 'photoframe', 'doorjamb', 'mouse', 'cupboard', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Turn on the water faucet. Wet my hands and turn off the faucet. Put soap on my hands. Rub hands together, making sure the soap covers all areas of the hands. Scrub hands for about 20 seconds. Turn on the water water faucet. Rinse hands under the running water. Turn off the water faucet. Dry hands with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'milk', 'bathroom_counter', 'floor', 'bookshelf', 'table_cloth', 'toaster', 'tvstand', 'shower', 'foundation', 'pillow', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'cup', 'coffe_maker', 'food_onion', 'kitchen_counter', 'dishrack', 'cupboard', 'clothes_dress', 'sink', 'bag', 'table', 'dishwasher', 'dirt', 'shoe_rack']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <dishrack>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <dishrack> <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'knife', 'drawing', 'wallshelf', 'orchid', 'hanger', 'cpuscreen', 'floor', 'character', 'keyboard', 'purse', 'homework', 'mousepad', 'wall', 'pasta', 'bench', 'keys', 'television', 'wooden_spoon', 'towel_rack', 'kettle', 'kitchen_counter', 'hair', 'conditioner', 'home_office', 'table']\nTask:Hang up car keys\nresponse:\n", "output": {"thoughts": "Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <purse>", "[FIND] <purse>", "[OPEN] <purse>", "[FIND] <keys>", "[GRAB] <keys>", "[CLOSE] <purse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'juice', 'microwave', 'chair', 'freezer', 'detergent', 'drawing', 'orchid', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'shower', 'mat', 'ceilinglamp', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'newspaper', 'coffe_maker', 'television', 'ice', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'cupboard', 'table', 'dishwasher', 'creditcard']\nTask:Grab some juice\nresponse:\n{\n    \"thoughts\": I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'microwave', 'chair', 'freezer', 'dog', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'phone', 'dvd_player', 'character', 'bookshelf', 'computer', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'bench', 'photoframe', 'food_food', 'tray', 'curtain', 'kitchen_counter', 'door', 'sink', 'table']\nTask:Pet dog\nresponse:\n", "output": {"thoughts": "Pet my dog in hall", "actions": ["[WALK] <home_office>", "[WALK] <dog>", "[TURNTO] <dog>", "[LOOKAT] <dog>", "[TOUCH] <dog>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'beer', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'food_orange', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Make toast\nresponse:\n{\n    \"thoughts\": Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[CLOSE] <freezer>', '[FIND] <toaster>', '[PLUGIN] <toaster>', '[PUTBACK] <food_bread> <toaster>', '[SWITCHON] <toaster>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'filing_cabinet', 'microwave', 'chair', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'face', 'toaster', 'stovefan', 'faucet', 'oil', 'towel', 'wall', 'facial_cleanser', 'check', 'coffe_maker', 'tray', 'rag', 'mouse', 'kitchen_counter', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Turn the faucet on. Wet the face. Take facial cleanser out of packaging. Lather on face. Rinse face with water. Turn off the faucet. Dry face with towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <face>", "[WASH] <face>", "[FIND] <rag>", "[GRAB] <rag>", "[RINSE] <rag>", "[FIND] <facial_cleanser>", "[GRAB] <facial_cleanser>", "[POUR] <facial_cleanser> <rag>", "[PUTOBJBACK] <facial_cleanser>", "[SCRUB] <face>", "[RINSE] <rag>", "[PUTOBJBACK] <rag>", "[RINSE] <face>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <face>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'beer', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'stovefan', 'mat', 'toaster', 'shower', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'food_food', 'food_donut', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Walk into the kitchen. Open the fridge. Take out the chicken. Open the kitchen cabinet. Take out the frying pan. Put the chicken in the frying pan. Put the frying pan on the stove. Turn on the stove. Cook the chicken.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[CLOSE] <freezer>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[CLOSE] <kitchen_cabinet>', '[WALK] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'ceilinglamp', 'tvstand', 'window', 'homework', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'box', 'cup', 'photoframe', 'television', 'curtain', 'doorjamb', 'dresser', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Straighten pictures on wall\nresponse:\n", "output": {"thoughts": "Move the pictures/paintings until they are straight.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'tablelamp', 'chair', 'walllamp', 'trashcan', 'knifeblock', 'bathroom_counter', 'keyboard', 'computer', 'brush', 'tvstand', 'shower', 'faucet', 'mousepad', 'wall', 'bench', 'cup', 'bed', 'photoframe', 'clothes_gloves', 'tray', 'wooden_spoon', 'towel_rack', 'dresser', 'cupboard', 'door', 'bathtub', 'napkin', 'bathroom_cabinet']\nTask:Wash monitor\nresponse:\n{\n    \"thoughts\": Clean the grime off of my monitor screen when not in use,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHOFF] <computer>', '[FIND] <brush>', '[TURNTO] <brush>', '[LOOKAT] <brush>', '[GRAB] <brush>', '[WIPE] <computer>', '[DROP] <brush>', '[FIND] <napkin>', '[GRAB] <napkin>', '[WASH] <napkin>', '[WIPE] <computer>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'shower', 'window', 'mousepad', 'dining_room', 'wall', 'check', 'television', 'wine', 'doorjamb', 'shaving_cream', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Complete surveys on amazon turk\nresponse:\n", "output": {"thoughts": "Sit down at computer, Turn computer on. Go online to Amazon mechanical turk, Browse through surveys, Select a survey.", "actions": ["[WALK] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[POINTAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>", "[PUTOBJBACK] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'oil', 'couch', 'wall', 'cup', 'food_food', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'console', 'hairbrush', 'cupboard', 'food_steak', 'maindoor', 'door', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch favorite shows, netflix, hulu,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'food_cake', 'mat', 'ceilinglamp', 'tvstand', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'wall_clock', 'photoframe', 'clothes_gloves', 'bed', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'home_office', 'table', 'light']\nTask:Admire art\nresponse:\n", "output": {"thoughts": "Looking at painting, art", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'toilet', 'detergent', 'orchid', 'longboard', 'floor', 'desk', 'food_bacon', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'balanceball', 'window', 'bed', 'oven_mitts', 'ice', 'curtain', 'after_shave', 'lighter', 'towel_rack', 'cards', 'controller', 'laundry_detergent', 'maindoor', 'door', 'sink', 'home_office', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Twist the lamp switch until the light goes on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'freezer', 'walllamp', 'trashcan', 'woman', 'bathroom_counter', 'ceiling', 'phone', 'bedroom', 'man', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'couch', 'dining_room', 'child', 'powersocket', 'bench', 'oven', 'photoframe', 'bed', 'coffe_maker', 'doorjamb', 'towel_rack', 'bathtub', 'sink', 'bathroom_cabinet', 'table']\nTask:Say goodbye to guests leaving\nresponse:\n", "output": {"thoughts": "Tell them bye", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[FIND] <man>", "[TOUCH] <man>", "[FIND] <woman>", "[TOUCH] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'tea_bag', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'bench', 'photoframe', 'television', 'tray', 'doorjamb', 'mouse', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Find computer. Sit in chair in front of computer desk. Turn on computer. Open internet browser. Browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'purse', 'mail', 'computer', 'laptop', 'deck_of_cards', 'faucet', 'foundation', 'mousepad', 'wall', 'wall_clock', 'bench', 'cup', 'doorjamb', 'mouse', 'door', 'light', 'table']\nTask:Sent email\nresponse:\n", "output": {"thoughts": "Walk into home office. Sit down in a chair at a desk. Turn on laptop. Bring up the internet. Go to gmail website. Begin composing an email. Hit the send button.", "actions": ["[WALK] <chair>", "[SIT] <chair>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <mail>", "[GRAB] <mail>", "[READ] <mail>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'food_bread', 'microwave', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'bathroom_counter', 'bedroom', 'tea_bag', 'tvstand', 'window', 'form', 'pillow', 'broom', 'bench', 'basket_for_clothes', 'newspaper', 'bed', 'check', 'coffe_maker', 'tray', 'door', 'home_office', 'light', 'bathroom_cabinet', 'tooth_paste', 'coffee']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": I get dressed in my bedroom in the mornings.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'walllamp', 'toothbrush', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'bench', 'box', 'cup', 'photoframe', 'novel', 'mouthwash', 'television', 'curtain', 'doorjamb', 'comb', 'light', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit and read", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[CLOSE] <filing_cabinet>", "[WALK] <couch>", "[SIT] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'detergent', 'knifeblock', 'phone', 'bookshelf', 'computer', 'mop_bucket', 'brush', 'clothes_socks', 'mousepad', 'dining_room', 'nightstand', 'basket_for_clothes', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'bathtub', 'sink', 'home_office', 'napkin', 'dishwasher', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'toilet', 'chair', 'freezer', 'bathroom', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'computer', 'laptop', 'tvstand', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'clothes_pants', 'wall', 'powersocket', 'bed', 'tray', 'doorjamb', 'diary', 'dresser', 'clothes_dress', 'sink', 'clothes_shirt', 'dirt']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "Find dresser, grab pants, grab shirt, put on pants, put on shirt", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'sponge', 'tablelamp', 'mop', 'bathroom', 'drawing', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'stamp', 'wine_glass', 'clothes_socks', 'couch', 'pillow', 'wall', 'nightstand', 'food_cheese', 'bed', 'photoframe', 'food_food', 'check', 'doorjamb', 'door', 'sink', 'bag', 'bathroom_cabinet', 'table', 'light']\nTask:Entertain\nresponse:\n{\n    \"thoughts\": Give guests wine, cheese, crackers and sit and talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <food_cheese>', '[FIND] <food_cheese>', '[FIND] <food_food>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <wine_glass>', '[GRAB] <wine_glass>', '[DRINK] <wine_glass>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'knife', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'character', 'ceilinglamp', 'mat', 'shower', 'food_jam', 'dining_room', 'wall', 'bench', 'cutting_board', 'food_vegetable', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'table', 'light']\nTask:Chop vegetables\nresponse:\n", "output": {"thoughts": "Cut up veggies with knife", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <cutting_board>", "[GRAB] <cutting_board>", "[CLOSE] <kitchen_cabinet>", "[WALK] <table>", "[PUTBACK] <knife> <table>", "[PUTBACK] <cutting_board> <table>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <food_vegetable> <cutting_board>", "[GRAB] <knife>", "[CUT] <food_vegetable>", "[PUTOBJBACK] <knife>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'homework', 'window', 'wall', 'powersocket', 'oven', 'toy', 'bed', 'toilet_paper', 'television', 'toothbrush_holder', 'food_donut', 'tray', 'doorjamb', 'door', 'sink', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Lift the remote and change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'food_butter', 'trashcan', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'mail', 'stovefan', 'mat', 'tvstand', 'stereo', 'window', 'wall', 'nightstand', 'child', 'oven', 'powersocket', 'bed', 'novel', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'home_office', 'table']\nTask:Story reading time\nresponse:\n", "output": {"thoughts": "Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[SIT] <bed>", "[TURNTO] <child>", "[LOOKAT] <child>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'detergent', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'coffee_filter', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'sink', 'bathtub', 'light', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I pick up remote. I turn on tv. I push on the remote. I change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'table']\nTask:Watch youtube\nresponse:\n", "output": {"thoughts": "Go to youtube.Com watch videos", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'brush', 'ceilinglamp', 'tvstand', 'plate', 'mousepad', 'closetdrawer', 'dustpan', 'pillow', 'wall', 'powersocket', 'television', 'tray', 'doorjamb', 'centerpiece', 'mouse', 'shoes', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": walk to living room, look at computer, find chair, sit in chair, grab keyboard, grab mouse, type using keyboard and mouse,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[WALK] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'mousepad', 'couch', 'towel', 'wall', 'bed', 'photoframe', 'curtain', 'after_shave', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'light', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap and hot water", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'milk', 'floor', 'ceiling', 'shampoo', 'purse', 'ceilinglamp', 'mat', 'faucet', 'oil', 'wall', 'pasta', 'tape', 'novel', 'food_food', 'food_donut', 'wooden_spoon', 'mouse', 'kitchen_counter', 'door', 'home_office', 'bathroom_cabinet', 'vase', 'soap', 'instrument_violin', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['juice', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'alcohol', 'faucet', 'closetdrawer', 'food_jam', 'wall', 'oven', 'cup', 'photoframe', 'coffe_maker', 'razor', 'curtain', 'doorjamb', 'towel_rack', 'sink', 'bathroom_cabinet', 'light', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "go to bathroom. walk to sink,turn on faucet, put hands in water. rub hands together. grab soap. rub soap in hands. put soap down. put hands under water. rubs hands together. turn off faucet", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'curtain', 'shoes', 'cd', 'vacuum_cleaner', 'dresser', 'cupboard', 'clothes_dress', 'home_office', 'table', 'shoe_rack']\nTask:Take off shoes\nresponse:\n{\n    \"thoughts\": Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'walllamp', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'mousepad', 'wall', 'bench', 'photoframe', 'novel', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'shaving_cream', 'door', 'bathroom_cabinet', 'light', 'home_office']\nTask:Read\nresponse:\n", "output": {"thoughts": "We pick out a book from the bookshelf and read it.", "actions": ["[WALK] <novel>", "[WALK] <bookshelf>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'trashcan', 'knifeblock', 'floor', 'character', 'computer', 'table_cloth', 'toaster', 'plate', 'faucet', 'dish_soap', 'wall', 'oven', 'bench', 'cup', 'coffe_maker', 'food_food', 'television', 'wine', 'tray', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['dog', 'freezer', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'computer', 'stovefan', 'tvstand', 'faucet', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'curtain', 'rag', 'doorjamb', 'towel_rack', 'sink', 'bathroom_cabinet', 'table', 'light']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'food_bacon', 'mat', 'ceilinglamp', 'shower', 'wall', 'wall_clock', 'cup', 'photoframe', 'bed', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'coffee']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[TURNTO] <cup>', '[LOOKAT] <cup>', '[WALK] <cup>', '[FIND] <freezer>', '[TURNTO] <freezer>', '[LOOKAT] <freezer>', '[WALK] <freezer>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'alcohol', 'toaster', 'tvstand', 'shower', 'closetdrawer', 'nightstand', 'wall_clock', 'bench', 'powersocket', 'food_food', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'sink', 'home_office', 'table', 'light', 'coffee']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'cat', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'mail', 'alcohol', 'mat', 'faucet', 'clothes_pants', 'pillow', 'wall', 'wall_clock', 'oven', 'pasta', 'bench', 'coffee_filter', 'ice', 'paper_towel', 'conditioner', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk to the bathroom and find the sink. I grab the soap and rub some on my hands. I turn the faucet on and put my hands in the water. I rinse my hands and then turn off the faucet. Then I dry my hands on a paper towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <hands_both>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'keyboard', 'mat', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'cupboard', 'door', 'light', 'table']\nTask:Gaze out window\nresponse:\n", "output": {"thoughts": "Look at the yard from the window", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[FIND] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'microwave', 'freezer', 'drawing', 'phone', 'bookshelf', 'toaster', 'stovefan', 'tvstand', 'mousepad', 'couch', 'clothes_hat', 'food_orange', 'wall', 'photoframe', 'clothes_gloves', 'bed', 'check', 'television', 'food_rice', 'tray', 'doorjamb', 'dresser', 'hairdryer', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'door']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to my bedroom, and then I switch on the radio,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'milk', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'mat', 'window', 'mousepad', 'closetdrawer', 'couch', 'wall', 'nightstand', 'bed', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'home_office', 'light']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Lay on pillows in bed. Turn on laptop. Go to internet on laptop and go to Netflix. Pick a television show on Netflix and hit play. Watch Netflix.", "actions": ["[WALK] <bedroom>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <bed>", "[SIT] <bed>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[PUTBACK] <laptop> <bed>", "[LIE] <bed>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'mat', 'shower', 'window', 'food_kiwi', 'wall', 'bench', 'newspaper', 'food_food', 'coffe_maker', 'food_onion', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'home_office']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'walllamp', 'cat', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'faucet', 'window', 'wall', 'nightstand', 'child', 'oven', 'bench', 'bed', 'novel', 'check', 'curtain', 'doorjamb', 'coffee_pot', 'towel_rack', 'door']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Read story to children", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <child>", "[GREET] <child>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[CLOSE] <filing_cabinet>", "[WALK] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'knifeblock', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'mat', 'tvstand', 'faucet', 'couch', 'pillow', 'wall', 'powersocket', 'bed', 'clothes_gloves', 'novel', 'toilet_paper', 'doorjamb', 'coffee_pot', 'mouse', 'kitchen_counter', 'door', 'napkin', 'bathroom_cabinet', 'light']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Hit the switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['toilet', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'face', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'wall_clock', 'face_soap', 'food_food', 'coffe_maker', 'tray', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'shoe_rack']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Run warm water. Cleanse my face with soap. Rinse my face with water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <face_soap>", "[GRAB] <face_soap>", "[FIND] <face>", "[SCRUB] <face>", "[PUTOBJBACK] <face_soap>", "[WASH] <face>", "[RINSE] <face>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bills', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'television', 'toothbrush_holder', 'wine', 'doorjamb', 'mouse', 'dresser', 'hairbrush', 'controller', 'clothes_dress', 'home_office', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'trashcan', 'drawing', 'needle', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'bed', 'newspaper', 'television', 'curtain', 'kitchen_counter', 'headset', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on the couch and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'brush', 'stovefan', 'faucet', 'iron', 'couch', 'towel', 'wall', 'powersocket', 'cutting_board', 'photoframe', 'bed', 'television', 'lighter', 'console', 'dresser', 'pajamas', 'door', 'sink', 'bathroom_cabinet', 'table', 'clothes_skirt', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'bathroom', 'hanger', 'bathroom_counter', 'man', 'bookshelf', 'keyboard', 'piano_bench', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'window', 'closetdrawer', 'dining_room', 'photoframe', 'tray', 'curtain', 'food_onion', 'mouse', 'dresser', 'cupboard', 'door', 'sink', 'home_office', 'light', 'clothes_skirt', 'bathroom_cabinet']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "When someone knocks I n the door walk through the hallway and greet them.", "actions": ["[RUN] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'tea', 'mop', 'microwave', 'toilet', 'toothbrush', 'drawing', 'knifeblock', 'orchid', 'phone', 'ceiling', 'computer', 'stovefan', 'toaster', 'tvstand', 'window', 'floor_lamp', 'couch', 'clothes_hat', 'photoframe', 'after_shave', 'food_onion', 'headset', 'chessboard', 'cupboard', 'controller', 'door', 'bathroom_cabinet', 'clothes_skirt']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'microwave', 'chair', 'walllamp', 'trashcan', 'milk', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'folder', 'desk', 'bedroom', 'alcohol', 'computer', 'mat', 'window', 'closetdrawer', 'candle', 'wall_clock', 'powersocket', 'oven_mitts', 'food_onion', 'shoes', 'door', 'sink', 'table']\nTask:Shut off alarm\nresponse:\n", "output": {"thoughts": "Shutting of the alarm clock", "actions": ["[WALK] <bedroom>", "[WALK] <alarm_clock>", "[FIND] <alarm_clock>", "[GRAB] <alarm_clock>", "[SWITCHOFF] <alarm_clock>", "[PUTOBJBACK] <alarm_clock>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'piano_bench', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'powersocket', 'doorjamb', 'mouse', 'dresser', 'maindoor', 'light', 'instrument_violin', 'shoe_rack']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Compose an email in the home office.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'chair', 'walllamp', 'bathroom', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'oven', 'newspaper', 'bed', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Use computer\nresponse:\n", "output": {"thoughts": "Sit at desk, use the computer mouse", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <mousepad>", "[PUTBACK] <mouse> <mousepad>", "[TURNTO] <computer>", "[POINTAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'measuring_cup', 'microwave', 'freezer', 'ceilingfan', 'milk', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'button', 'toaster', 'faucet', 'picture', 'wall', 'nightstand', 'oven', 'photoframe', 'clothes_gloves', 'coffe_maker', 'food_food', 'curtain', 'centerpiece', 'food_sugar', 'vacuum_cleaner', 'chessboard', 'coffee_cup', 'dishwasher', 'light']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[LOOKAT] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <milk>', '[GRAB] <milk>', '[PUTBACK] <milk> <coffe_maker>', '[FIND] <food_sugar>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[SWITCHON] <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cat', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'television', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'sink', 'bathroom_cabinet', 'table']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Press the power button on the computer to turn it on", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'microwave', 'freezer', 'ground_coffee', 'hanger', 'floor', 'ceiling', 'stamp', 'keyboard', 'toaster', 'plate', 'faucet', 'shower', 'mousepad', 'food_kiwi', 'wall', 'wall_clock', 'oven', 'cutting_board', 'standingmirror', 'photoframe', 'coffe_maker', 'video_game_controller', 'razor', 'coffee_filter', 'water', 'coffee_pot', 'kitchen_counter', 'sink', 'home_office']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'orchid', 'dvd_player', 'desk', 'bedroom', 'food_cake', 'face', 'alcohol', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'bed', 'check', 'food_food', 'coffe_maker', 'mirror', 'kitchen_counter', 'dishrack', 'hair', 'clothes_dress', 'sink', 'table']\nTask:Check appearance in mirror\nresponse:\n", "output": {"thoughts": "Walk to full-length mirror. Look in mirror. Make sure everything about your appearance is adequate. Adjust clothes as needed.", "actions": ["[WALK] <bedroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>", "[FIND] <face>", "[TURNTO] <face>", "[LOOKAT] <face>", "[FIND] <hair>", "[TURNTO] <hair>", "[LOOKAT] <hair>", "[FIND] <clothes_dress>", "[TURNTO] <clothes_dress>", "[LOOKAT] <clothes_dress>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'freezer', 'food_carrot', 'milk', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'mousepad', 'garbage_can', 'clothes_hat', 'food_orange', 'food_dessert', 'wall', 'powersocket', 'food_vegetable', 'food_cheese', 'doorjamb', 'lighter', 'mouse', 'coffee', 'door', 'light', 'colander', 'teeth']\nTask:Rotate stock in refrigerator\nresponse:\n{\n    \"thoughts\": Throw out expired items, put older items in front.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[TURNTO] <food_bread>', '[LOOKAT] <food_bread>', '[PUTOBJBACK] <food_bread>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[FIND] <garbage_can>', '[PUTBACK] <food_cheese> <garbage_can>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[PUTOBJBACK] <food_cheese>', '[FIND] <food_dessert>', '[TURNTO] <food_dessert>', '[LOOKAT] <food_dessert>', '[PUSH] <food_dessert>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <milk>', '[GRAB] <milk>', '[TURNTO] <milk>', '[LOOKAT] <milk>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'ground_coffee', 'trashcan', 'drawing', 'floor', 'character', 'bedroom', 'stovefan', 'mat', 'shower', 'faucet', 'floor_lamp', 'window', 'clothes_hat', 'wall', 'nightstand', 'powersocket', 'wall_clock', 'cup', 'bed', 'curtain', 'console', 'mouse', 'kitchen_counter', 'cupboard', 'bathtub', 'home_office', 'light']\nTask:Turn on lights\nresponse:\n", "output": {"thoughts": "Walk to your lamp, switch it on, and lay on your bed", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'jelly', 'chair', 'freezer', 'bathroom', 'food_carrot', 'knifeblock', 'bathroom_counter', 'keyboard', 'alcohol', 'clothes_jacket', 'stovefan', 'mat', 'toaster', 'stereo', 'picture', 'ceilinglamp', 'dining_room', 'bowl', 'wall', 'oven', 'powersocket', 'cup', 'photoframe', 'check', 'video_game_controller', 'razor', 'food_food', 'curtain', 'shoes']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": turn radio on, listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tea', 'toilet', 'walllamp', 'bathroom', 'detergent', 'drawing', 'bills', 'floor', 'character', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'stereo', 'closetdrawer', 'form', 'dining_room', 'nightstand', 'powersocket', 'cup', 'bed', 'tray', 'mouse', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'home_office', 'dirt', 'electrical_outlet']\nTask:Turn on radio\nresponse:\n", "output": {"thoughts": "Go to the radio.  Switch on the power button.  Tune the station.  Set the volume.  Listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <electrical_outlet>", "[TURNTO] <electrical_outlet>", "[PLUGIN] <stereo>", "[SWITCHON] <electrical_outlet>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'filing_cabinet', 'tablelamp', 'chair', 'freezer', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'purse', 'computer', 'mail', 'toaster', 'tvstand', 'mousepad', 'wall', 'check', 'ice', 'tray', 'doorjamb', 'towel_rack', 'coffee_pot', 'cards', 'dresser', 'cupboard', 'door', 'napkin', 'light', 'scrabble']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk over to the phone. .I reach my arm out. I open my hand. I place my hand on the phone. I close my fingers around the phone. I pick up the phone.,\n    \"actions\": ['[WALK] <phone>', '[TOUCH] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'food_pizza', 'tablelamp', 'blender', 'food_salt', 'blow_dryer', 'trashcan', 'toothbrush', 'cpuscreen', 'hanger', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'purse', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'microphone', 'oven', 'food_food', 'check', 'tray', 'dresser', 'napkin', 'bathroom_cabinet', 'clothes_shirt']\nTask:Leave home\nresponse:\n", "output": {"thoughts": "Put on shoes. Put on coat. Open door and exit through it. Lock door behind you.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'newspaper', 'television', 'food_donut', 'food_rice', 'doorjamb', 'mouse', 'cd', 'dresser']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'faucet', 'closetdrawer', 'pillow', 'nightstand', 'bench', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'after_shave', 'mouse', 'cupboard', 'door', 'bathtub', 'sink', 'light', 'clothes_shirt']\nTask:Open door\nresponse:\n", "output": {"thoughts": "Open the door when leaving or coming in.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'food_butter', 'drawing', 'cpuscreen', 'hanger', 'floor', 'phone', 'desk', 'stamp', 'keyboard', 'computer', 'alcohol', 'toaster', 'laptop', 'mat', 'mousepad', 'wall', 'oven', 'bed', 'mouthwash', 'doorjamb', 'mouse', 'chessboard', 'door', 'clothes_scarf']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": walk to home office and switch on laptop and with the useof mouse and keyboard send mail,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'drawing', 'hanger', 'floor', 'ceiling', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'foundation', 'oil', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'novel', 'television', 'curtain', 'after_shave', 'door', 'napkin', 'home_office', 'table']\nTask:Turn off TV\nresponse:\n", "output": {"thoughts": "Walk over to tv and press power button", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'table_cloth', 'stovefan', 'mat', 'faucet', 'foundation', 'mousepad', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'toy', 'hands_both', 'tray', 'doorjamb', 'towel_rack', 'shaving_cream', 'door', 'sink', 'spoon', 'bathroom_cabinet', 'light', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I need to rinse my dentures off and put them back in my mouth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <teeth>', '[GRAB] <teeth>', '[WASH] <teeth>', '[PUTOBJBACK] <teeth>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <towel>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['food_bread', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'couch', 'wall', 'nightstand', 'oven', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'food_onion', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'electrical_outlet']\nTask:Toast bread\nresponse:\n", "output": {"thoughts": "Place two slices of bread in the toaster. Select medium toast on toaster.  Push down button to begin toasting.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[WALK] <electrical_outlet>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <electrical_outlet>", "[TURNTO] <toaster>", "[LOOKAT] <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'toilet', 'chair', 'blow_dryer', 'hanger', 'bathroom_counter', 'ceiling', 'phone', 'desk', 'mail', 'shoe-shine_kit', 'mousepad', 'food_noodles', 'cutting_board', 'cup', 'photoframe', 'bed', 'coffe_maker', 'television', 'curtain', 'centerpiece', 'comb', 'pajamas', 'clothes_dress', 'door', 'home_office', 'light', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into the living room, sit down by the cat, put out hand, stroke the cat from head to tail,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'freezer', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'bookshelf', 'ceilinglamp', 'tvstand', 'shower', 'mat', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'oven', 'coffe_maker', 'tray', 'curtain', 'coffee_pot', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'hairdryer']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Walk up to the door.  Open the door and greet whomever is out side", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'bathroom', 'cat', 'knifeblock', 'drawing', 'wallshelf', 'phone', 'ceiling', 'character', 'keyboard', 'toaster', 'mat', 'tvstand', 'electric_shaver', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'photoframe', 'novel', 'food_food', 'curtain', 'doorjamb', 'dishrack', 'pajamas', 'door', 'bathtub', 'spoon', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to phone, grab phone, pick up phone, put to ear,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'wallshelf', 'bathroom_counter', 'ceiling', 'desk', 'bookshelf', 'food_bacon', 'faucet', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'toilet_paper', 'check', 'razor', 'curtain', 'doorjamb', 'towel_rack', 'chessboard', 'shoes', 'cupboard', 'sink', 'bathroom_cabinet']\nTask:Get toilet paper\nresponse:\n", "output": {"thoughts": "There are extra toilet papers in the bottom cabinet.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <toilet_paper>", "[GRAB] <toilet_paper>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'walllamp', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'shampoo', 'dvd_player', 'character', 'keyboard', 'mop_bucket', 'toaster', 'dining_room', 'wall', 'oven', 'cutting_board', 'light_bulb', 'coffe_maker', 'food_food', 'razor', 'tray', 'wooden_spoon', 'pencil', 'dishrack', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'colander', 'laser_pointer']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <dishrack>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <dishrack> <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'microwave', 'toilet', 'freezer', 'trashcan', 'food_carrot', 'drawing', 'floor', 'ceiling', 'mat', 'ceilinglamp', 'shower', 'couch', 'wall', 'novel', 'food_food', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'home_office', 'instrument_guitar', 'bathroom_cabinet', 'table', 'dirt']\nTask:Added meat to freezer\nresponse:\n", "output": {"thoughts": "After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I go to the home office to find the computer. I turn switch computer on , i open the email. i write the email and  send it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'tea', 'walllamp', 'cat', 'trashcan', 'food_butter', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'shower', 'closetdrawer', 'couch', 'dining_room', 'oven', 'bench', 'bed', 'food_food', 'television', 'food_onion', 'doorjamb', 'cupboard', 'sink', 'bathtub', 'soap', 'laser_pointer']\nTask:Walk to room\nresponse:\n", "output": {"thoughts": "Leave the living room and enter the hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'cat', 'cpuscreen', 'desk', 'bookshelf', 'computer', 'alcohol', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'stereo', 'shower', 'closetdrawer', 'candle', 'basket_for_clothes', 'photoframe', 'food_food', 'check', 'television', 'water', 'curtain', 'lighter', 'doorjamb', 'vacuum_cleaner', 'cupboard', 'home_office', 'table', 'bathroom_cabinet']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'bills', 'cpuscreen', 'bathroom_counter', 'bedroom', 'computer', 'mail', 'bookmark', 'shower', 'couch', 'towel', 'pillow', 'wall_clock', 'oven', 'powersocket', 'bench', 'bed', 'photoframe', 'television', 'mirror', 'curtain', 'towel_rack', 'kitchen_counter', 'home_office', 'dirt']\nTask:Look at mirror\nresponse:\n", "output": {"thoughts": "I look at my reflection to check my cleanliness.", "actions": ["[WALK] <bathroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'ground_coffee', 'bathroom', 'detergent', 'dough', 'knifeblock', 'drawing', 'food_egg', 'floor', 'bedroom', 'keyboard', 'purse', 'mop_bucket', 'stovefan', 'mousepad', 'food_orange', 'bed', 'check', 'toilet_paper', 'food_rice', 'tray', 'shaving_cream', 'mouse', 'cd', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'clothes_skirt', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Twist the lamp switch until the light goes on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'toaster', 'window', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'wall_clock', 'television', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'clothes_shirt']\nTask:Organize closet\nresponse:\n", "output": {"thoughts": "Organize the hanging clothes in my closet by color and type and get rid of clothes I don't wear.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PULL] <clothes_pants>", "[PUTIN] <clothes_pants> <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PULL] <clothes_pants>", "[PUTIN] <clothes_pants> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUSH] <clothes_shirt>", "[PUTIN] <clothes_shirt> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUSH] <clothes_shirt>", "[PUTIN] <clothes_shirt> <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[CLOSE] <dresser>", "[WALK] <basket_for_clothes>", "[PUTBACK] <clothes_shirt> <basket_for_clothes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'freezer', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'toaster', 'tvstand', 'faucet', 'mousepad', 'towel', 'wall', 'wall_clock', 'bench', 'powersocket', 'nightstand', 'bed', 'toilet_paper', 'television', 'wine', 'curtain', 'doorjamb', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'clothes_skirt', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. Walk to the sink. Turn on tap. Get soap. Wash and rinse hands. Turn off tap. Dry hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[LOOKAT] <sink>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[RINSE] <soap>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'phone', 'mat', 'ceilinglamp', 'shower', 'window', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'food_food', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'bathtub']\nTask:Look in refrigerator\nresponse:\n", "output": {"thoughts": "Approach the refrigerator. Use the handle to open the refrigerator door. Look at each shelf of the refrigerator. Look in each refrigerator drawer. Close the refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[WALK] <freezer>", "[OPEN] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <food_food>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'band-aids', 'trashcan', 'dough', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'dvd_player', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'mousepad', 'food_orange', 'wall', 'oven', 'keys', 'mouse', 'shoes', 'hair', 'dishwasher', 'vase', 'coffee']\nTask:Read book\nresponse:\n{\n    \"thoughts\": find a book to read. sit down. read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_pizza', 'toilet', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'toaster', 'shower', 'faucet', 'closetdrawer', 'clothes_hat', 'pillow', 'wall', 'wall_clock', 'nightstand', 'food_food', 'curtain', 'towel_rack', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'sink', 'home_office', 'light', 'bathroom_cabinet']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Get the vacuum and use it to clean my floors", "actions": ["[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[WALK] <bedroom>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'chair', 'freezer', 'ground_coffee', 'floor', 'ceiling', 'mail', 'food_bacon', 'toaster', 'deck_of_cards', 'faucet', 'window', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toy', 'light_bulb', 'coffe_maker', 'coffee_filter', 'water', 'curtain', 'after_shave', 'pencil', 'kitchen_counter', 'pajamas', 'sink', 'light']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[FIND] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'stovefan', 'toaster', 'mat', 'faucet', 'ceilinglamp', 'shower', 'pillow', 'wall', 'wall_clock', 'oven', 'bench', 'powersocket', 'tape', 'coffe_maker', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light']\nTask:Keep an eye on stove as something is cooking\nresponse:\n", "output": {"thoughts": "Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <pot>", "[TURNTO] <pot>", "[LOOKAT] <pot>", "[FIND] <wall_clock>", "[TURNTO] <wall_clock>", "[LOOKAT] <wall_clock>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'freezer', 'cat', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'oil', 'couch', 'clothes_hat', 'wall', 'basket_for_clothes', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'towel_rack', 'comb', 'dresser', 'maindoor', 'soap']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I put my groceries into the fridge.,\n    \"actions\": ['[WALK] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'knife', 'cpuscreen', 'bathroom_counter', 'floor', 'keyboard', 'face', 'ceilinglamp', 'mat', 'faucet', 'shower', 'mousepad', 'closetdrawer', 'towel', 'bench', 'oven', 'bed', 'coffe_maker', 'television', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'chessboard', 'hair', 'controller', 'bathtub', 'bathroom_cabinet']\nTask:Rain welcome\nresponse:\n", "output": {"thoughts": "The robot would greet me and then hand me a towel.  It would make an appropriate quip about the weather while I dried off.  It would receive the wet towel when I was done.", "actions": ["[WALK] <bedroom>", "[WALK] <towel>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <face>", "[WIPE] <face>", "[FIND] <hair>", "[WIPE] <hair>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'food_ice_cream', 'chair', 'bathroom', 'trashcan', 'food_carrot', 'shampoo', 'phone', 'floor', 'stamp', 'bookshelf', 'ceilinglamp', 'window', 'homework', 'couch', 'clothes_pants', 'pillow', 'nightstand', 'cup', 'bed', 'check', 'food_food', 'ice', 'headset', 'cupboard', 'door', 'home_office', 'light', 'dishwasher', 'clothes_shirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to phone, grab phone, pick up phone, put to ear,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'phone', 'ceiling', 'keyboard', 'computer', 'mat', 'tvstand', 'faucet', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'photoframe', 'tray', 'doorjamb', 'towel_rack', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table']\nTask:Close door\nresponse:\n", "output": {"thoughts": "Push the door shut until it latches.", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'walllamp', 'ground_coffee', 'orchid', 'drawing', 'cpuscreen', 'floor', 'phone', 'ceiling', 'toaster', 'ceilinglamp', 'plate', 'tvstand', 'window', 'dish_soap', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'razor', 'tray', 'mouse', 'kitchen_counter', 'cupboard', 'bathtub', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'milk', 'drawing', 'bathroom_counter', 'floor', 'phone', 'character', 'bedroom', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'check', 'television', 'curtain', 'mouse', 'door', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Do homework\nresponse:\n", "output": {"thoughts": "Sit on the couch in the Living room and pick u some papers", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <check>", "[GRAB] <check>", "[DROP] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'chair', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'toothbrush', 'needle', 'desk', 'keyboard', 'alcohol', 'tvstand', 'faucet', 'window', 'food_noodles', 'couch', 'form', 'pillow', 'wall', 'wall_clock', 'bench', 'bed', 'curtain', 'cupboard', 'instrument_violin', 'bathtub', 'home_office', 'light', 'board_game']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": i go find the phone. i pick up the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'woman', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'man', 'bookshelf', 'keyboard', 'piano_bench', 'stovefan', 'tvstand', 'shower', 'faucet', 'oil', 'dining_room', 'food_kiwi', 'nightstand', 'bench', 'oven', 'powersocket', 'coffe_maker', 'tray', 'curtain', 'towel_rack', 'cupboard', 'door', 'sink', 'light']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Stand next to the door. When the door opens from the outside, greet the guests.", "actions": ["[RUN] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'bathroom', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'purse', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'couch', 'dining_room', 'wall', 'bench', 'food_food', 'coffe_maker', 'television', 'toilet_paper', 'wooden_spoon', 'curtain', 'pencil', 'shoes', 'table', 'light', 'vase']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go the living room, look for my tv remote control, take the control and then sit down on the couch and turn on the TV and push the buttons on the remote control to change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'food_butter', 'drawing', 'needle', 'bathroom_counter', 'floor', 'character', 'bedroom', 'bookshelf', 'alcohol', 'computer', 'brush', 'tvstand', 'faucet', 'shower', 'foundation', 'window', 'mousepad', 'bench', 'photoframe', 'television', 'mirror', 'rag', 'towel_rack', 'mouse', 'bathtub', 'light', 'table', 'scrabble']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray glass cleaner on mirror. Wipe with a clean cloth.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[WALK] <mirror>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'chair', 'detergent', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'toaster', 'mat', 'plate', 'faucet', 'ceilinglamp', 'dish_soap', 'wall', 'food_cheese', 'light_bulb', 'novel', 'coffe_maker', 'food_food', 'television', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'shoes', 'dishrack', 'dresser', 'cupboard', 'sink', 'home_office']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Put all dirty dishes on one side. Take sponge and cover it in soap. Soap each dish. Rinse each dish with water. Put on drying rack.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[RINSE] <sponge>', '[SQUEEZE] <sponge>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <sponge>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'brush', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'wall', 'powersocket', 'bench', 'photoframe', 'bed', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'pencil', 'door', 'bathtub', 'napkin', 'home_office', 'table']\nTask:Wash monitor\nresponse:\n", "output": {"thoughts": "Clean the grime off of my monitor screen when not in use", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>", "[FIND] <brush>", "[TURNTO] <brush>", "[LOOKAT] <brush>", "[GRAB] <brush>", "[WIPE] <computer>", "[DROP] <brush>", "[FIND] <napkin>", "[GRAB] <napkin>", "[WASH] <napkin>", "[WIPE] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'crayon', 'freezer', 'bathroom', 'food_carrot', 'drawing', 'knifeblock', 'floor', 'desk', 'ceilinglamp', 'stereo', 'window', 'couch', 'food_orange', 'pillow', 'form', 'dining_room', 'wall', 'bench', 'coffe_maker', 'television', 'rag', 'pencil', 'comb', 'table', 'tooth_paste', 'clothes_shirt', 'scrabble', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'chair', 'bathroom', 'cat', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'clothes_underwear', 'phone', 'ceiling', 'dvd_player', 'desk', 'stovefan', 'faucet', 'shower', 'closetdrawer', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'bed', 'tray', 'mouse', 'dresser', 'home_office']\nTask:Keep cats out of room\nresponse:\n", "output": {"thoughts": "Stand in the doorway and create a loud noise so the cats will stay out.", "actions": ["[WALK] <bedroom>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'ceilinglamp', 'food_noodles', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'headset', 'dresser', 'food_steak', 'sink']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": change the tv channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'trashcan', 'drawing', 'duster', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'mat', 'tvstand', 'picture', 'window', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'powersocket', 'photoframe', 'food_food', 'curtain', 'doorjamb', 'sink', 'bathroom_cabinet', 'light', 'home_office', 'scrabble']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Using the duster, lightly dust all hanging pictures on walls in living room. Dust all around surface of picture frame and front of pictures, moving duster back and forth, side to side, until all dust is gone.", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <duster>", "[GRAB] <duster>", "[WALK] <wall>", "[WALK] <picture>", "[WIPE] <picture>", "[WALK] <picture>", "[WIPE] <picture>", "[WALK] <picture>", "[WIPE] <picture>", "[WALK] <filing_cabinet>", "[PUTBACK] <duster> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'stovefan', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'tape', 'newspaper', 'bed', 'television', 'doorjamb', 'mouse', 'sink', 'home_office', 'table', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office. Sit on chair. Switch on computer. type in keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'toaster', 'window', 'closetdrawer', 'couch', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'light_bulb', 'check', 'television', 'curtain', 'doorjamb', 'dresser', 'bathtub', 'table', 'light']\nTask:Straighten paintings on wall\nresponse:\n", "output": {"thoughts": "Go up to the painting and tilt it which ever way it is uneven until it looks straight.", "actions": ["[WALK] <home_office>", "[WALK] <wall>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'chair', 'freezer', 'walllamp', 'detergent', 'orchid', 'phone', 'bookshelf', 'computer', 'toaster', 'tvstand', 'stereo', 'closetdrawer', 'food_peanut_butter', 'clothes_pants', 'form', 'wall', 'oven', 'cup', 'bed', 'mouthwash', 'rag', 'towel_rack', 'kitchen_counter', 'pajamas', 'food_steak', 'door', 'napkin', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the playback device you want to listen to. turn on. press play. enjoyu music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'bed', 'coin', 'curtain', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'light']\nTask:Wipe down baseboards please\nresponse:\n", "output": {"thoughts": "Use a rag to get all the dust off the baseboards please.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <wall>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'faucet', 'picture', 'window', 'mousepad', 'oil', 'clothes_hat', 'wall', 'keys', 'food_food', 'coin', 'food_onion', 'mouse', 'cupboard', 'laundry_detergent', 'door', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I walk into the home office and walk to the desk. I pull out the chair and sit down. I turn on the computer. After the computer is turned on I click on my email program.  When the program is loaded I write an email.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'trashcan', 'bills', 'knife', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'keyboard', 'bookmark', 'faucet', 'window', 'closetdrawer', 'towel', 'dining_room', 'wall_clock', 'check', 'mouthwash', 'television', 'coffe_maker', 'rag', 'doorjamb', 'mouse', 'pencil', 'dishrack', 'sink', 'bathroom_cabinet', 'table', 'light']\nTask:Open front door\nresponse:\n", "output": {"thoughts": "Walk over to the front door, turn the knob, and open the door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sponge', 'filing_cabinet', 'tea', 'chair', 'walllamp', 'bathroom', 'ceilingfan', 'character', 'computer', 'food_bacon', 'mop_bucket', 'balanceball', 'foundation', 'form', 'bowl', 'wall', 'pasta', 'basket_for_clothes', 'bed', 'check', 'doorjamb', 'diary', 'hairbrush', 'cupboard', 'dresser', 'laundry_detergent', 'food_steak', 'light', 'scrabble']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'purse', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'shower', 'wall', 'powersocket', 'bench', 'cup', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'door', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Grasp light switch, flip light switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tea', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'pasta', 'powersocket', 'coffe_maker', 'doorjamb', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'home_office', 'table', 'tooth_paste']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Push power button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'desk', 'bookshelf', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'shower', 'electric_shaver', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'bed', 'television', 'oven_mitts', 'curtain', 'cupboard', 'bathtub', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Read magazine\nresponse:\n", "output": {"thoughts": "Please sit on the couch and read a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <newspaper>", "[LOOKAT] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'chair', 'freezer', 'drawing', 'orchid', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'ceilinglamp', 'plate', 'mat', 'stereo', 'shower', 'couch', 'pillow', 'food_kiwi', 'wall', 'pasta', 'bed', 'novel', 'food_food', 'check', 'television', 'console', 'pencil', 'kitchen_counter', 'cupboard', 'table']\nTask:Cut bread\nresponse:\n{\n    \"thoughts\": I get a slice of bread, grab a knife, and then cut the piece of bread in half.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <kitchen_cabinet>', '[FIND] <table>', '[PUTBACK] <plate> <table>', '[PUTBACK] <knife> <table>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[CLOSE] <freezer>', '[PUTBACK] <food_bread> <plate>', '[GRAB] <knife>', '[CUT] <food_bread>', '[PUTOBJBACK] <knife>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'ground_coffee', 'detergent', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'floor', 'bookshelf', 'computer', 'button', 'mat', 'ceilinglamp', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'wall_clock', 'basket_for_clothes', 'photoframe', 'toothbrush_holder', 'tray', 'doorjamb', 'towel_rack', 'vacuum_cleaner', 'dresser', 'home_office', 'light']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Robot has a vacuum underneath its feat/body that will allow it to clean the surface of the floor.", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[WALK] <dining_room>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[WALK] <home_office>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <filing_cabinet>", "[PUTBACK] <vacuum_cleaner> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'microwave', 'cd_player', 'freezer', 'band-aids', 'bathroom', 'food_carrot', 'orchid', 'music_stand', 'phone', 'bookshelf', 'thread', 'mail', 'stereo', 'homework', 'mousepad', 'closetdrawer', 'clothes_hat', 'form', 'broom', 'newspaper', 'photoframe', 'check', 'food_rice', 'food_donut', 'dresser', 'sink', 'napkin', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to the living and turn to the music stand. Switch on the CD player to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <music_stand>', '[FIND] <music_stand>', '[TURNTO] <music_stand>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'brush', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'bench', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'light']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Find the power button. Press.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'purse', 'toaster', 'faucet', 'shower', 'homework', 'pillow', 'wall', 'oven', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to kitchen, turn on water, find soap, apply to hand, lather, rinse, dry hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['food_bread', 'microwave', 'toilet', 'freezer', 'trashcan', 'milk', 'drawing', 'floor', 'ceiling', 'phone', 'keyboard', 'mat', 'tvstand', 'shower', 'oil', 'garbage_can', 'form', 'food_dessert', 'food_kiwi', 'wall', 'food_cheese', 'food_vegetable', 'doorjamb', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathroom_cabinet']\nTask:Rotate stock in refrigerator\nresponse:\n", "output": {"thoughts": "Throw out expired items, put older items in front.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[TURNTO] <food_bread>", "[LOOKAT] <food_bread>", "[PUTOBJBACK] <food_bread>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[FIND] <garbage_can>", "[PUTBACK] <food_cheese> <garbage_can>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[PUTOBJBACK] <food_cheese>", "[FIND] <food_dessert>", "[TURNTO] <food_dessert>", "[LOOKAT] <food_dessert>", "[PUSH] <food_dessert>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <milk>", "[GRAB] <milk>", "[TURNTO] <milk>", "[LOOKAT] <milk>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'sheets', 'filing_cabinet', 'food_bread', 'food_salt', 'microwave', 'chair', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'mail', 'stovefan', 'ceilinglamp', 'mat', 'stereo', 'couch', 'wall', 'powersocket', 'cup', 'check', 'mouthwash', 'food_donut', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'table', 'teeth']\nTask:Play games\nresponse:\n{\n    \"thoughts\": I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <board_game>', '[FIND] <board_game>', '[GRAB] <board_game>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <board_game> <table>']\n}\n", "input": "\nScene_Objects:['chair', 'ground_coffee', 'walllamp', 'drawing', 'floor', 'ceiling', 'computer', 'stovefan', 'toaster', 'tvstand', 'faucet', 'shower', 'closetdrawer', 'wall', 'oven', 'bench', 'coffe_maker', 'television', 'coffee_filter', 'water', 'tray', 'curtain', 'doorjamb', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub', 'napkin']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[GRAB] <ground_coffee>", "[PUTOBJBACK] <ground_coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'dry_pasta', 'crayon', 'band-aids', 'detergent', 'drawing', 'clothes_underwear', 'table_cloth', 'stovefan', 'laptop', 'toaster', 'faucet', 'ceilinglamp', 'closetdrawer', 'wall', 'nightstand', 'oven', 'keys', 'food_food', 'tray', 'coffee_pot', 'pencil', 'headset', 'door', 'hairdryer', 'sink', 'table', 'dishwasher', 'clothes_shirt', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": You walk to the living room and find the cat. You look at the cat and touch it to pet it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'toilet', 'chair', 'bathroom', 'drawing', 'orchid', 'bathroom_counter', 'phone', 'character', 'bedroom', 'toaster', 'mat', 'stereo', 'faucet', 'floor_lamp', 'shower', 'couch', 'dining_room', 'pasta', 'oven', 'powersocket', 'love_seat', 'newspaper', 'photoframe', 'coffe_maker', 'tray', 'dresser', 'bathtub', 'soap']\nTask:Read\nresponse:\n", "output": {"thoughts": "Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'computer', 'mat', 'shoe-shine_kit', 'oil', 'pillow', 'wall', 'pasta', 'powersocket', 'cup', 'food_food', 'video_game_controller', 'toothbrush_holder', 'television', 'doorjamb', 'kettle', 'door', 'light', 'clothes_skirt', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on light in an area,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['fork', 'sheets', 'microwave', 'toilet', 'freezer', 'trashcan', 'food_carrot', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'ceilinglamp', 'shower', 'oil', 'pillow', 'wall', 'novel', 'food_food', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub', 'sink', 'home_office', 'light']\nTask:Put away groceries\nresponse:\n", "output": {"thoughts": "Place groceries in refrigerator from shopping at the grocery store.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'dough', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'window', 'mousepad', 'food_noodles', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'pencil', 'hair', 'door', 'table', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to the living room and sit on the couch. Turn on television using remote control. Use remote control buttons to navigate to desired channel. Watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mop_bucket', 'mat', 'laptop', 'window', 'homework', 'mousepad', 'food_orange', 'dining_room', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'bed', 'tray', 'curtain', 'mouse', 'pencil', 'light']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Bring laptop to bed and watch a movie", "actions": ["[WALK] <bedroom>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[WALK] <bed>", "[SIT] <bed>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'computer', 'stovefan', 'faucet', 'window', 'mousepad', 'closetdrawer', 'couch', 'towel', 'wall', 'nightstand', 'curtain', 'towel_rack', 'mouse', 'coffee_pot', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom and locate the sink. I turn the faucet so that water is running, and rinse my hands briefly. Then I grab the soap and squeeze some onto my hands. I spend a few minutes rubbing my hands together, then I rinse them again. Finally, I turn the faucet off and wipe my hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SQUEEZE] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'walllamp', 'bathroom', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'newspaper', 'bed', 'check', 'television', 'doorjamb', 'towel_rack', 'mouse', 'bathtub', 'table']\nTask:Research\nresponse:\n", "output": {"thoughts": "I read different articles online to gain a better understanding of a particular business, industry, method, terminology, etc.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'purse', 'stovefan', 'ceilinglamp', 'stereo', 'faucet', 'window', 'balanceball', 'electric_shaver', 'couch', 'wall', 'bench', 'food_food', 'television', 'curtain', 'diary', 'mouse', 'dresser', 'hair', 'bathroom_cabinet', 'table', 'home_office', 'vase', 'teeth', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'bills', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'food_food', 'television', 'doorjamb', 'towel_rack', 'mouse', 'bathroom_cabinet', 'table', 'home_office']\nTask:Work\nresponse:\n", "output": {"thoughts": "Sit at a desk.  Turn on computer.  Sign onto mTurk.  Perform tasks.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'mop', 'walllamp', 'trashcan', 'wallshelf', 'floor', 'ceiling', 'folder', 'bedroom', 'tvstand', 'shower', 'foundation', 'window', 'mousepad', 'wall', 'powersocket', 'photoframe', 'light_bulb', 'food_food', 'coffe_maker', 'television', 'food_donut', 'curtain', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to the bathroom and wash with the shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>', '[WASH] <shower>']\n}\n", "input": "\nScene_Objects:['juice', 'microwave', 'toilet', 'freezer', 'trashcan', 'food_carrot', 'milk', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'computer', 'mail', 'mat', 'shower', 'pillow', 'dining_room', 'wall', 'nightstand', 'doorjamb', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'table', 'tooth_paste']\nTask:Get glass of milk\nresponse:\n", "output": {"thoughts": "Get an empty glass. Take milk from refrigerator and open it. Pour milk into glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'deck_of_cards', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'door', 'light', 'table']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Keep my keyboard vacuumed and free of debris.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'walllamp', 'trashcan', 'drawing', 'knife', 'floor', 'ceiling', 'bedroom', 'keyboard', 'mat', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'food_food', 'television', 'tray', 'curtain', 'console', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'table', 'board_game']\nTask:Pull up carpet\nresponse:\n", "output": {"thoughts": "Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.", "actions": ["[WALK] <home_office>", "[WALK] <mat>", "[FIND] <mat>", "[TOUCH] <mat>", "[PULL] <mat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink']\nTask:Use computer\nresponse:\n{\n    \"thoughts\": Browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'trashcan', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'purse', 'mail', 'toaster', 'tvstand', 'faucet', 'towel', 'wall', 'nightstand', 'powersocket', 'food_food', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Turn faucet clockwise to turn on water.  Place hands under stream of water, pick up soap, rub hands with soap.  Put soap back down.  Rub hands together with soap on them.  Rinse hands under water from faucet.  Turn faucet counterclockwise to turn water off.  Dry hands by rubbing them on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'toilet', 'freezer', 'nail_polish', 'detergent', 'longboard', 'floor', 'ceiling', 'bedroom', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'food_peanut_butter', 'form', 'wall', 'nightstand', 'oven', 'food_food', 'ice', 'after_shave', 'doorjamb', 'lighter', 'coffee_pot', 'mouse', 'spoon']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Put groceries into the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'shampoo', 'character', 'wine_glass', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'couch', 'pillow', 'wall', 'bench', 'powersocket', 'food_cheese', 'food_food', 'coffe_maker', 'television', 'wine', 'curtain', 'table']\nTask:Entertain\nresponse:\n", "output": {"thoughts": "Give guests wine, cheese, crackers and sit and talk", "actions": ["[WALK] <home_office>", "[WALK] <food_cheese>", "[FIND] <food_cheese>", "[FIND] <food_food>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <wine_glass>", "[GRAB] <wine_glass>", "[DRINK] <wine_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'bathroom', 'detergent', 'wallshelf', 'hanger', 'floor', 'character', 'keyboard', 'envelope', 'clothes_jacket', 'stovefan', 'stereo', 'window', 'food_peanut_butter', 'nightstand', 'clothes_gloves', 'food_food', 'food_donut', 'after_shave', 'centerpiece', 'towel_rack', 'cupboard', 'door', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'table', 'home_office']\nTask:Put on coat\nresponse:\n{\n    \"thoughts\": I remove my scarf from a coat peg in the entrance hall.  I wrap the scarf around my neck,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <hanger>', '[TURNTO] <hanger>', '[LOOKAT] <hanger>', '[FIND] <clothes_scarf>', '[GRAB] <clothes_scarf>', '[PUTON] <clothes_scarf>']\n}\n", "input": "\nScene_Objects:['food_bread', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'window', 'couch', 'pillow', 'wall', 'oven', 'coffe_maker', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'food_steak', 'door', 'sink', 'table']\nTask:Make toast\nresponse:\n", "output": {"thoughts": "Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'folder', 'tea_bag', 'mat', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'pasta', 'nightstand', 'photoframe', 'novel', 'food_food', 'tray', 'curtain', 'doorjamb', 'door', 'table', 'light', 'board_game']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Bring book, lie on the sofa, read book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[LIE] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['jelly', 'bathroom', 'trashcan', 'wallshelf', 'clothes_underwear', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'faucet', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall_clock', 'newspaper', 'coffe_maker', 'television', 'curtain', 'centerpiece', 'towel_rack', 'kitchen_counter', 'cupboard', 'bathtub', 'sink', 'clothes_scarf', 'light']\nTask:Watch fly\nresponse:\n", "output": {"thoughts": "Look around for a fly, often on a wall or ceiling.  Follow it's movements.", "actions": ["[WALK] <bedroom>", "[WALK] <fly>", "[FIND] <fly>", "[TURNTO] <fly>", "[LOOKAT] <fly>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_bread', 'bathroom', 'trashcan', 'food_carrot', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'phone', 'folder', 'bedroom', 'desk', 'keyboard', 'toaster', 'bench', 'tray', 'doorjamb', 'diary', 'console', 'mouse', 'towel_rack', 'cd', 'dresser', 'cupboard', 'controller', 'conditioner', 'table', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to table, grab water glass and drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>', '[PUTBACK] <water_glass> <table>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'bathroom', 'drawing', 'floor', 'phone', 'stamp', 'keyboard', 'computer', 'toaster', 'mat', 'tvstand', 'faucet', 'window', 'couch', 'wall', 'nightstand', 'powersocket', 'bed', 'clothes_gloves', 'novel', 'coffe_maker', 'food_donut', 'tray', 'curtain', 'doorjamb', 'cupboard', 'door', 'sink', 'bathtub', 'home_office']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I lie on my bed and read a book.", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'bathroom', 'trashcan', 'ceilingfan', 'bills', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'folder', 'stovefan', 'tvstand', 'window', 'homework', 'wall', 'oven', 'powersocket', 'cutting_board', 'basket_for_clothes', 'food_cheese', 'clothes_gloves', 'check', 'television', 'toothbrush_holder', 'curtain', 'bathroom_cabinet', 'laser_pointer']\nTask:Open window\nresponse:\n{\n    \"thoughts\": Unlock the latch at the top. Push the bottom half of the window upwards.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <window>', '[OPEN] <window>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'character', 'bookshelf', 'keyboard', 'tvstand', 'shower', 'window', 'closetdrawer', 'wall', 'wall_clock', 'food_cheese', 'photoframe', 'check', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'shredder', 'bathtub', 'sink', 'light', 'table', 'dirt', 'electrical_outlet']\nTask:Shred receipts\nresponse:\n", "output": {"thoughts": "Take old receipts and put through the shredder", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[FIND] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <check>", "[GRAB] <check>", "[CLOSE] <filing_cabinet>", "[FIND] <electrical_outlet>", "[FIND] <shredder>", "[PLUGIN] <shredder>", "[SWITCHON] <shredder>", "[PUTBACK] <check> <shredder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'food_ice_cream', 'walllamp', 'nail_polish', 'detergent', 'knife', 'needle', 'cpuscreen', 'bathroom_counter', 'character', 'bedroom', 'remote_control', 'envelope', 'alcohol', 'button', 'deck_of_cards', 'iron', 'nightstand', 'pasta', 'television', 'wine', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher', 'vase']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'freezer', 'trashcan', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'window', 'closetdrawer', 'clothes_pants', 'wall', 'bed', 'coffe_maker', 'television', 'tray', 'towel_rack', 'dresser', 'cupboard', 'hair', 'bathtub', 'sink', 'home_office', 'table', 'bathroom_cabinet', 'clothes_shirt']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "Get some clothes together that make an appropriate outfit.  Put them on", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTON] <clothes_pants>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'bedroom', 'computer', 'alcohol', 'mat', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'bench', 'cutting_board', 'standingmirror', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'chessboard', 'dresser', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit on the couch. Turn on TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'jelly', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'homework', 'closetdrawer', 'couch', 'pillow', 'wall', 'newspaper', 'coffe_maker', 'television', 'tray', 'curtain', 'towel_rack', 'cupboard', 'home_office', 'table']\nTask:Read newspaper\nresponse:\n", "output": {"thoughts": "Read the newspaper on the couch", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'faucet', 'clothes_socks', 'microphone', 'wall', 'bench', 'powersocket', 'photoframe', 'food_rice', 'food_onion', 'towel_rack', 'cd', 'hairbrush', 'controller', 'maindoor', 'sink', 'bathroom_cabinet', 'home_office', 'clothes_skirt', 'vase', 'soap', 'instrument_violin', 'laser_pointer', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to bathroom, go to sink, turn the water on via faucet, grab soap, use the soap to scrub my hands, put the soap back and wash my hands. Then I stop the water with faucet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'food_carrot', 'woman', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'child', 'bed', 'television', 'curtain', 'towel_rack', 'dresser', 'table']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sitting and watching  a movie with family", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TURNTO] <child>", "[LOOKAT] <child>", "[GREET] <child>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'microwave', 'walllamp', 'drawing', 'food_egg', 'wallshelf', 'floor', 'stamp', 'computer', 'alcohol', 'toaster', 'plate', 'shower', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'water', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'spoon', 'instrument_guitar', 'light', 'dishwasher', 'soap', 'scrabble']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": first i will take all the dishes to the dish washer and with the help of dishwasher and water i will clean all the dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <water>', '[TURNTO] <water>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['crayon', 'chair', 'walllamp', 'drawing', 'floor', 'ceiling', 'video_game_console', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'video_game_controller', 'television', 'curtain', 'after_shave', 'mouse', 'kitchen_counter', 'bathtub', 'table', 'soap']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on game.  Use controller", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[PLUGIN] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'measuring_cup', 'ground_coffee', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'keyboard', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'oven', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'toy', 'television', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'table']\nTask:Pay bills\nresponse:\n", "output": {"thoughts": "Check my bills, on the computer, open the appropriate sites, pay my due bills", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'remote_control', 'mop_bucket', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'cup', 'newspaper', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'shoes', 'dresser', 'door', 'home_office', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'ceiling', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'child', 'bench', 'coffe_maker', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'home_office', 'table', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn on tv.  Sit with kids.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[FIND] <child>", "[GREET] <child>", "[FIND] <child>", "[GREET] <child>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>", "[TURNTO] <television>", "[WATCH] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'toilet', 'chair', 'bathroom', 'orchid', 'knifeblock', 'floor', 'phone', 'desk', 'keyboard', 'food_cake', 'foundation', 'floor_lamp', 'pillow', 'nightstand', 'powersocket', 'box', 'basket_for_clothes', 'razor', 'tray', 'doorjamb', 'cupboard', 'bathtub', 'spoon', 'home_office', 'table', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to living room, find lamp, switch on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['microwave', 'bathroom', 'detergent', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'video_game_console', 'keyboard', 'stovefan', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall_clock', 'oven', 'bed', 'coffe_maker', 'video_game_controller', 'television', 'doorjamb', 'cd', 'cupboard', 'door', 'bathtub', 'bathroom_cabinet', 'light', 'home_office', 'table']\nTask:Playing video game\nresponse:\n", "output": {"thoughts": "Playing  video games  on various video game  systems such as wii u  and  xbox", "actions": ["[WALK] <home_office>", "[WALK] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[PUTOBJBACK] <video_game_controller>", "[SWITCHOFF] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'microwave', 'chair', 'freezer', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'stereo', 'picture', 'mat', 'closetdrawer', 'couch', 'dining_room', 'wall_clock', 'bench', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'home_office', 'instrument_guitar']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the living room and then I walk to the light switch in there and turn the lights on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['juice', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'floor', 'ceiling', 'desk', 'mop_bucket', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'wall', 'wall_clock', 'oven', 'nightstand', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'food_snack', 'door', 'sink', 'bathtub']\nTask:Have snack\nresponse:\n", "output": {"thoughts": "Pour soda into a glass. Get pretzels out of bag. Eat and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTIN] <juice> <freezer>", "[CLOSE] <freezer>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <food_snack>", "[GRAB] <food_snack>", "[CLOSE] <cupboard>", "[DRINK] <cup>", "[EAT] <food_snack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'dough', 'wallshelf', 'hanger', 'floor', 'phone', 'folder', 'desk', 'tea_bag', 'mousepad', 'closetdrawer', 'couch', 'dustpan', 'pillow', 'bed', 'television', 'toothbrush_holder', 'food_onion', 'curtain', 'lighter', 'diary', 'cd', 'dresser', 'controller', 'sink', 'table', 'light', 'dishwasher']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to home office 1. Walk over to phone 1 on table 1.  Grab phone 1.  lift phone 1.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'microwave', 'toilet', 'freezer', 'orchid', 'wallshelf', 'hanger', 'floor', 'stovefan', 'toaster', 'faucet', 'mousepad', 'closetdrawer', 'form', 'dining_room', 'toilet_paper', 'television', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'bathtub', 'spoon', 'bathroom_cabinet', 'table', 'light', 'instrument_violin', 'laser_pointer', 'teeth']\nTask:Open door\nresponse:\n", "output": {"thoughts": "I walk up to the door.  I use my hand to turn the door knob to the left.  I pull the door open with the same hand I used to turn the door knob", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'dog', 'toothbrush', 'knifeblock', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'character', 'desk', 'mat', 'tvstand', 'shower', 'homework', 'dining_room', 'wall', 'bed', 'clothes_gloves', 'toothbrush_holder', 'after_shave', 'doorjamb', 'mouse', 'pencil', 'cd', 'tooth_paste', 'shoe_rack']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'tablelamp', 'bathroom', 'trashcan', 'food_butter', 'cpuscreen', 'desk', 'bookshelf', 'computer', 'mop_bucket', 'mat', 'closetdrawer', 'form', 'dining_room', 'nightstand', 'bench', 'oven', 'powersocket', 'toy', 'photoframe', 'coffe_maker', 'television', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'sink', 'home_office']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Grab toys off floor.  Put it in a container.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'mop', 'toilet', 'chair', 'walllamp', 'bathroom', 'food_carrot', 'hanger', 'floor', 'shampoo', 'bookshelf', 'purse', 'computer', 'stovefan', 'toaster', 'shower', 'mousepad', 'dustpan', 'clothes_pants', 'wall', 'cutting_board', 'standingmirror', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'hairbrush', 'door', 'bathtub']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom. Pull down pants. Use the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <clothes_pants>', '[FIND] <clothes_pants>', '[PULL] <clothes_pants>', '[FIND] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'freezer', 'walllamp', 'trashcan', 'woman', 'toothbrush', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'character', 'desk', 'thread', 'stovefan', 'mat', 'faucet', 'window', 'mousepad', 'couch', 'bench', 'bed', 'tray', 'door', 'bathtub', 'bathroom_cabinet', 'table']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet people into my home", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'freezer', 'walllamp', 'bathroom', 'bills', 'needle', 'drawing', 'character', 'table_cloth', 'toaster', 'picture', 'floor_lamp', 'window', 'mousepad', 'microphone', 'food_kiwi', 'wall', 'wall_clock', 'newspaper', 'bed', 'television', 'tray', 'cupboard', 'controller', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[TOUCH] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'walllamp', 'food_butter', 'cpuscreen', 'bedroom', 'bookshelf', 'toaster', 'mat', 'deck_of_cards', 'shower', 'brush', 'window', 'clothes_hat', 'pillow', 'dining_room', 'wall_clock', 'bench', 'toy', 'rag', 'doorjamb', 'mouse', 'pencil', 'door', 'bathroom_cabinet', 'home_office', 'light']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Enter inside kid's room, begin picking up toys from the floor and putting them in a toy box, leave once once.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'conditioner', 'door', 'table']\nTask:Check email\nresponse:\n{\n    \"thoughts\": I log onto my computer and then go to the e-mail app, I click on each new email and read them,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mail>', '[GRAB] <mail>', '[READ] <mail>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'cat', 'drawing', 'floor', 'dvd_player', 'bookshelf', 'food_cake', 'keyboard', 'computer', 'bookmark', 'mat', 'shower', 'faucet', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'television', 'curtain', 'towel_rack', 'mouse', 'door', 'sink', 'bathtub', 'light']\nTask:Breakfast\nresponse:\n", "output": {"thoughts": "I have breakfast in bed, watching tv.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[GRAB] <food_food>", "[WALK] <bedroom>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <bed>", "[LIE] <bed>", "[TURNTO] <television>", "[WATCH] <television>", "[EAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'tablelamp', 'chair', 'dog', 'walllamp', 'ceilingfan', 'knifeblock', 'needle', 'bathroom_counter', 'floor', 'computer', 'shoe-shine_kit', 'dustpan', 'box', 'cup', 'bed', 'keys', 'television', 'food_donut', 'tray', 'curtain', 'doorjamb', 'kettle', 'kitchen_counter', 'shoes', 'comb', 'sink', 'bathroom_cabinet', 'board_game']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": go to the bathroom and brush your teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'dog', 'trashcan', 'drawing', 'cpuscreen', 'character', 'keyboard', 'bookshelf', 'toaster', 'mat', 'tvstand', 'faucet', 'window', 'mousepad', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'wooden_spoon', 'kettle', 'shaving_cream', 'towel_rack', 'cupboard', 'conditioner', 'bathtub', 'dirt']\nTask:Lock door\nresponse:\n", "output": {"thoughts": "Turn the lock.", "actions": ["[WALK] <bedroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'microwave', 'chair', 'wallshelf', 'hanger', 'floor', 'phone', 'food_bacon', 'piano_bench', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'closetdrawer', 'powersocket', 'television', 'coin', 'food_donut', 'ice', 'rag', 'pencil', 'shoes', 'dresser', 'clothes_dress', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Hang up jacket\nresponse:\n{\n    \"thoughts\": I put clothes on hangers and hang them in the closet,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[FIND] <hanger>', '[OPEN] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'keyboard', 'computer', 'toaster', 'tvstand', 'floor_lamp', 'window', 'closetdrawer', 'couch', 'towel', 'wall', 'wall_clock', 'food_food', 'doorjamb', 'mouse', 'headset', 'shoes', 'dresser', 'controller', 'bathtub', 'sink', 'home_office', 'light']\nTask:Turn night light on\nresponse:\n", "output": {"thoughts": "I walk to the dresser and in the middle Io to small battery operated lamp and switch it on", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'ground_coffee', 'walllamp', 'detergent', 'trashcan', 'drawing', 'wallshelf', 'character', 'remote_control', 'button', 'shower', 'faucet', 'window', 'pillow', 'wall', 'bench', 'newspaper', 'coffee_table', 'photoframe', 'coffe_maker', 'television', 'food_donut', 'tray', 'sink', 'napkin', 'bathroom_cabinet', 'home_office', 'instrument_guitar', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'button', 'clothes_jacket', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'powersocket', 'bed', 'check', 'food_food', 'food_onion', 'dresser', 'light']\nTask:Put on glasses\nresponse:\n", "output": {"thoughts": "Put on glasses.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[FIND] <dresser>", "[FIND] <spectacles>", "[GRAB] <spectacles>", "[PUTON] <spectacles>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'chair', 'freezer', 'drawing', 'floor', 'ceiling', 'character', 'toaster', 'plate', 'shower', 'floor_lamp', 'window', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'oven', 'toilet_paper', 'curtain', 'doorjamb', 'towel_rack', 'shoes', 'cd', 'laundry_detergent', 'door', 'light', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to wall, find light switch, turn on lights,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[WALK] <bedroom>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'microwave', 'freezer', 'walllamp', 'bathroom', 'orchid', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'computer', 'mat', 'tvstand', 'faucet', 'window', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'food_cheese', 'bed', 'novel', 'curtain', 'diary', 'dresser', 'bathroom_cabinet']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Put child in bed, sit down,read book.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'toilet', 'chair', 'freezer', 'wallshelf', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'toaster', 'ceilinglamp', 'window', 'pillow', 'wall', 'powersocket', 'oven', 'bench', 'food_cheese', 'bed', 'photoframe', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'lighter', 'kitchen_counter', 'cupboard', 'home_office']\nTask:Eat cheese\nresponse:\n{\n    \"thoughts\": Open fridge.  Get cheese.  Eat it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[EAT] <food_cheese>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'orchid', 'bills', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'desk', 'bookshelf', 'keyboard', 'table_cloth', 'piano_bench', 'ceilinglamp', 'tvstand', 'shower', 'window', 'couch', 'dining_room', 'wall', 'bench', 'photoframe', 'check', 'coffe_maker', 'food_food', 'doorjamb', 'dresser', 'bathtub', 'sink', 'light']\nTask:File documents\nresponse:\n", "output": {"thoughts": "I file all of my paperwork (including bills, receipts, documents, device instructions, etc.) in my black paperwork case that has many different files in order to organize.  I place each piece of paper in its' corresponding subject (as indicated by printed words on each tab).", "actions": ["[WALK] <home_office>", "[WALK] <bills>", "[FIND] <bills>", "[GRAB] <bills>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTIN] <bills> <filing_cabinet>", "[FIND] <check>", "[GRAB] <check>", "[WALK] <filing_cabinet>", "[PUTIN] <check> <filing_cabinet>", "[FIND] <check>", "[GRAB] <check>", "[WALK] <filing_cabinet>", "[PUTIN] <check> <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'cup', 'photoframe', 'television', 'tray', 'after_shave', 'doorjamb', 'coffee_pot', 'dresser', 'door', 'scrabble']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": watching TV from sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[DROP] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'tea_bag', 'computer', 'food_bacon', 'tvstand', 'iron', 'window', 'electric_shaver', 'mousepad', 'dining_room', 'wall', 'nightstand', 'wall_clock', 'photoframe', 'towel_rack', 'mouse', 'table']\nTask:Check email\nresponse:\n", "output": {"thoughts": "Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[WALK] <chair>", "[SIT] <chair>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'beer', 'walllamp', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'purse', 'food_bacon', 'piano_bench', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'shower', 'window', 'foundation', 'couch', 'pillow', 'wall', 'bench', 'check', 'wine', 'curtain', 'hairbrush', 'controller', 'door', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'floor', 'bedroom', 'bookshelf', 'keyboard', 'computer', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'novel', 'check', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'light', 'table']\nTask:Read\nresponse:\n", "output": {"thoughts": "Read a book in the living room", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[CLOSE] <filing_cabinet>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'scissors', 'tablelamp', 'crayon', 'microwave', 'freezer', 'bathroom', 'trashcan', 'ceilingfan', 'knifeblock', 'clothes_underwear', 'longboard', 'shampoo', 'dvd_player', 'folder', 'mail', 'bookmark', 'toaster', 'shower', 'homework', 'couch', 'bed', 'clothes_gloves', 'towel_rack', 'mouse', 'pencil', 'cupboard', 'maindoor', 'table', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'television', 'coin', 'doorjamb', 'mouse', 'chessboard', 'bathtub', 'table']\nTask:Write  school paper\nresponse:\n", "output": {"thoughts": "Writing  and drafting a paper for  class work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'freezer', 'bathroom', 'orchid', 'phone', 'dvd_player', 'character', 'bedroom', 'ceilinglamp', 'balanceball', 'iron', 'window', 'couch', 'food_orange', 'wall', 'tape', 'cup', 'basket_for_clothes', 'check', 'wooden_spoon', 'console', 'mouse', 'kitchen_counter', 'hairbrush', 'food_steak', 'sink', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'drawing', 'bills', 'cpuscreen', 'wallshelf', 'floor', 'desk', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'bench', 'bed', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'door', 'home_office', 'table', 'light']\nTask:Relax\nresponse:\n", "output": {"thoughts": "Lie down on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_ice_cream', 'walllamp', 'nail_polish', 'cat', 'orchid', 'drawing', 'floor', 'ceiling', 'purse', 'alcohol', 'mail', 'table_cloth', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'towel', 'wall', 'oven', 'cup', 'food_cheese', 'television', 'curtain', 'doorjamb', 'door', 'table', 'clothes_shirt']\nTask:Straighten paintings on wall\nresponse:\n{\n    \"thoughts\": Go up to the painting and tilt it which ever way it is uneven until it looks straight.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <wall>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TOUCH] <drawing>', '[PULL] <drawing>', '[PUSH] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'dining_room', 'wall', 'powersocket', 'oven', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Push power button.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'bathroom', 'drawing', 'floor', 'character', 'bedroom', 'stamp', 'computer', 'mat', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'closetdrawer', 'clothes_hat', 'pillow', 'dining_room', 'broom', 'wall', 'bed', 'doorjamb', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'home_office', 'board_game', 'dirt']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to living room ,find electrical outlet , switch on eletrical outlet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <electrical_outlet>', '[FIND] <electrical_outlet>', '[SWITCHON] <electrical_outlet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'walllamp', 'nail_polish', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'character', 'bookshelf', 'alcohol', 'toaster', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'bench', 'food_food', 'rag', 'mouse', 'kitchen_counter', 'door', 'sink', 'napkin', 'bathroom_cabinet', 'instrument_guitar']\nTask:Wash sink\nresponse:\n", "output": {"thoughts": "I spray some cleaner and wait a minte or so then wipe it out.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[WALK] <sink>", "[POUR] <cleaning_solution> <sink>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[WIPE] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'beer', 'freezer', 'walllamp', 'bathroom', 'bathroom_counter', 'character', 'bedroom', 'desk', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'clothes_pants', 'towel', 'bench', 'basket_for_clothes', 'bed', 'coin', 'food_donut', 'curtain', 'towel_rack', 'mouse', 'cd', 'comb', 'laundry_detergent', 'bathtub', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'microwave', 'toilet', 'freezer', 'ceiling', 'clothes_jacket', 'table_cloth', 'toaster', 'stovefan', 'plate', 'faucet', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'table']\nTask:Get out dish\nresponse:\n", "output": {"thoughts": "Open cupboard.  Get out the dish.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'bathroom', 'ceilingfan', 'milk', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'shoe-shine_kit', 'couch', 'food_jam', 'microphone', 'wall', 'nightstand', 'oven', 'cutting_board', 'television', 'curtain', 'pencil', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": We sit all around the television and watch a program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['crayon', 'trashcan', 'orchid', 'cpuscreen', 'ceiling', 'bedroom', 'table_cloth', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'powersocket', 'bench', 'bed', 'novel', 'food_food', 'mouthwash', 'coffe_maker', 'oven_mitts', 'curtain', 'doorjamb', 'cards', 'dresser', 'cupboard', 'bathtub', 'home_office', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the book shelf, get a book, open the book and read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'chair', 'cat', 'drawing', 'floor', 'tea_bag', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'stovefan', 'window', 'couch', 'food_orange', 'wall', 'food_food', 'television', 'oven_mitts', 'food_onion', 'curtain', 'after_shave', 'comb', 'napkin', 'home_office', 'table', 'dishwasher', 'bathroom_cabinet', 'laser_pointer', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room,find the sofa,sat down in the sofa,take some relaxation.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'wallshelf', 'floor', 'phone', 'character', 'bedroom', 'alcohol', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'foundation', 'food_noodles', 'wall', 'nightstand', 'bench', 'oven', 'food_food', 'television', 'curtain', 'rag', 'pencil', 'door', 'clothes_scarf', 'home_office', 'table']\nTask:Play games\nresponse:\n", "output": {"thoughts": "I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.", "actions": ["[WALK] <home_office>", "[WALK] <board_game>", "[FIND] <board_game>", "[GRAB] <board_game>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <board_game> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'detergent', 'milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'napkin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Find the remote, turn on the TV, watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_cake', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'couch', 'wall', 'powersocket', 'bed', 'television', 'doorjamb', 'mouse', 'hairdryer', 'door', 'sink', 'home_office', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Using desktop computer in home office", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[WALK] <desk>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'newspaper', 'television', 'food_donut', 'food_rice', 'doorjamb', 'mouse', 'cd', 'dresser']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'cd_player', 'chair', 'food_carrot', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'remote_control', 'bookshelf', 'computer', 'mat', 'shower', 'mousepad', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'coffe_maker', 'tray', 'towel_rack', 'mouse', 'coffee_pot', 'dresser', 'cupboard', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.", "actions": ["[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'scissors', 'tablelamp', 'freezer', 'trashcan', 'drawing', 'floor', 'shampoo', 'folder', 'bookshelf', 'purse', 'computer', 'window', 'wall', 'powersocket', 'cutting_board', 'cup', 'bed', 'photoframe', 'television', 'curtain', 'food_onion', 'shaving_cream', 'mouse', 'pencil', 'hair', 'maindoor', 'clothes_scarf', 'home_office', 'colander']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go t the bedroom and go to sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'freezer', 'walllamp', 'trashcan', 'dough', 'drawing', 'wallshelf', 'floor', 'phone', 'character', 'alcohol', 'mat', 'tvstand', 'ceilinglamp', 'laptop', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'wall_clock', 'powersocket', 'cup', 'photoframe', 'television', 'curtain', 'mouse', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Find couch. Sit on couch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'mop', 'toilet', 'dog', 'freezer', 'longboard', 'floor', 'ceiling', 'purse', 'ceilinglamp', 'balanceball', 'faucet', 'foundation', 'food_jam', 'wall', 'oven', 'cup', 'bed', 'novel', 'food_food', 'coffe_maker', 'coffee_filter', 'curtain', 'console', 'pencil', 'maindoor', 'table', 'light', 'dirt']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": put away groceries in fridge. open the door, put on shelf, close the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'bookmark', 'toaster', 'mat', 'tvstand', 'faucet', 'shower', 'laptop', 'mousepad', 'wall', 'oven', 'powersocket', 'towel_rack', 'kitchen_counter', 'headset', 'door', 'sink', 'bathtub', 'home_office', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I go to the bathroom. I locate the toilet. I walk to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[WALK] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'milk', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'mat', 'window', 'mousepad', 'closetdrawer', 'couch', 'wall', 'nightstand', 'bed', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'home_office', 'light']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Lay on pillows in bed. Turn on laptop. Go to internet on laptop and go to Netflix. Pick a television show on Netflix and hit play. Watch Netflix.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <bed>', '[SIT] <bed>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[PUTBACK] <laptop> <bed>', '[LIE] <bed>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['microwave', 'nail_polish', 'detergent', 'orchid', 'drawing', 'hanger', 'floor', 'phone', 'desk', 'button', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'dining_room', 'nightstand', 'wall_clock', 'oven', 'cup', 'photoframe', 'bed', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into the living room and turn on the radio.  Then, I select the station I want, which is usually a preset.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'faucet', 'picture', 'window', 'dustpan', 'towel', 'wall', 'nightstand', 'photoframe', 'keys', 'toothbrush_holder', 'food_donut', 'mouse', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Cleaning hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'dog', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'piano_bench', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'pillow', 'towel', 'wall', 'oven', 'coffe_maker', 'television', 'coffee_filter', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[PUTBACK] <coffee> <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'walllamp', 'ground_coffee', 'trashcan', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'shower', 'foundation', 'couch', 'microphone', 'wall', 'powersocket', 'oven', 'bed', 'television', 'wooden_spoon', 'doorjamb', 'lighter', 'dresser', 'door', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'television', 'curtain', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to Living room. Find Television. Walk to sofa. Sit on sofa. Find Remote. Grab remote. Turn to television. Look at television. Switch on Television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'piano_bench', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'powersocket', 'doorjamb', 'mouse', 'dresser', 'maindoor', 'light', 'instrument_violin', 'shoe_rack']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Compose an email in the home office.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'electric_shaver', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'coffe_maker', 'television', 'tray', 'doorjamb', 'mouse', 'bathtub', 'sink', 'home_office', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office, walk to desk, sit in chair, turn on computer, look at computer, click mouse and type on keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'shoe_rack', 'food_salt', 'mop', 'microwave', 'dog', 'jelly', 'walllamp', 'drawing', 'needle', 'food_egg', 'desk', 'stamp', 'computer', 'ceilinglamp', 'window', 'mousepad', 'food_orange', 'pillow', 'bench', 'cup', 'oven_mitts', 'wooden_spoon', 'pencil', 'dresser', 'maindoor', 'bag', 'table', 'instrument_violin', 'creditcard']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": go to radio. run on radio.,\n    \"actions\": ['[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'stereo', 'window', 'shower', 'mousepad', 'wall', 'powersocket', 'bench', 'clothes_gloves', 'television', 'tray', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'bathtub', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk into home office. I walk to desk. I sit at chair. I turn on computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'remote_control', 'computer', 'button', 'toaster', 'mat', 'shower', 'food_peanut_butter', 'food_kiwi', 'nightstand', 'oven', 'photoframe', 'television', 'oven_mitts', 'tray', 'kettle', 'cards', 'pencil', 'kitchen_counter', 'towel_rack', 'bathtub', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk into living room. Find television remote control. Pick up remote control. Press button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'character', 'shoe-shine_kit', 'faucet', 'couch', 'form', 'wall', 'bench', 'toy', 'cup', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom, find sink, turn on faucet, put soap in hands, wash and rinse hands,", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <soap>", "[WASH] <soap>", "[RINSE] <soap>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'measuring_cup', 'toilet', 'drawing', 'needle', 'hanger', 'floor', 'computer', 'mat', 'mousepad', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'bed', 'novel', 'food_food', 'video_game_controller', 'coffe_maker', 'tray', 'food_onion', 'doorjamb', 'lighter', 'cd', 'clothes_dress', 'bathroom_cabinet', 'light', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I would go lie down in my bed and open the book and start reading.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'microwave', 'toilet', 'chair', 'trashcan', 'drawing', 'wallshelf', 'bedroom', 'desk', 'keyboard', 'stovefan', 'laptop', 'toaster', 'window', 'closetdrawer', 'couch', 'broom', 'wall', 'powersocket', 'bench', 'photoframe', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'cupboard', 'door']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to bedroom, sit down, pick up book, read", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'tea', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'dresser', 'door', 'home_office']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Keep my keyboard vacuumed and free of debris.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'remote_control', 'computer', 'button', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'wall_clock', 'television', 'curtain', 'towel_rack', 'mouse', 'dresser', 'bathtub', 'bathroom_cabinet', 'table', 'coffee']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "change the tv channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blow_dryer', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'remote_control', 'thread', 'computer', 'alcohol', 'bookmark', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'tape', 'standingmirror', 'bed', 'television', 'curtain', 'doorjamb', 'instrument_guitar', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Changing channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[DROP] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'toaster', 'ceilinglamp', 'faucet', 'pillow', 'wall', 'bench', 'bed', 'coffe_maker', 'food_food', 'television', 'mouse', 'shoes', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'soap', 'clothes_shirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash Hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <faucet>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'ceilingfan', 'food_egg', 'hanger', 'ceiling', 'dvd_player', 'character', 'thread', 'clothes_jacket', 'toaster', 'deck_of_cards', 'foundation', 'mousepad', 'oil', 'pillow', 'oven', 'toy', 'food_food', 'check', 'lighter', 'centerpiece', 'dishrack', 'coffee', 'hair', 'laundry_detergent', 'maindoor', 'sink', 'door', 'teeth']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into my bedroom, go to the lamp and use the switch to turn it on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[TURNTO] <floor_lamp>', '[LOOKAT] <floor_lamp>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'bathroom', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'toaster', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'nightstand', 'television', 'toothbrush_holder', 'curtain', 'mouse', 'shoes', 'clothes_dress', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "go to bathroom. grab toothbrush. put toothpaste on toothbrush. brush teeth. turn on faucet. rinse toothbrush and mouth. put toothbrush away.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'jelly', 'bathroom', 'detergent', 'dough', 'orchid', 'floor', 'bedroom', 'mat', 'dish_soap', 'mousepad', 'dining_room', 'wall', 'nightstand', 'oven', 'food_cheese', 'photoframe', 'coffe_maker', 'food_food', 'ice', 'tray', 'kitchen_counter', 'headset', 'dishrack', 'dresser', 'cupboard', 'door', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the dishwasher and put dish soap. I close the dishwasher and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'bathroom', 'trashcan', 'food_butter', 'drawing', 'floor', 'ceiling', 'phone', 'mat', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'food_food', 'check', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'door', 'sink', 'spoon', 'home_office', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Do homework\nresponse:\n{\n    \"thoughts\": I do my online homework on the computer.  I use Pearson My Math Lab for math class and excel for my engineering class.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['blow_dryer', 'cat', 'needle', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'computer', 'mop_bucket', 'mat', 'faucet', 'mousepad', 'teeth', 'towel', 'wall', 'nightstand', 'bench', 'novel', 'doorjamb', 'towel_rack', 'kitchen_counter', 'laundry_detergent', 'door', 'sink', 'bathroom_cabinet', 'light', 'instrument_violin', 'hands_both']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I need to rinse my dentures off and put them back in my mouth.", "actions": ["[WALK] <bathroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <teeth>", "[GRAB] <teeth>", "[WASH] <teeth>", "[PUTOBJBACK] <teeth>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <towel>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'juice', 'nail_polish', 'wallshelf', 'floor', 'ceiling', 'face', 'mat', 'iron', 'shower', 'stereo', 'clothes_socks', 'food_jam', 'wall_clock', 'light_bulb', 'food_food', 'razor', 'mirror', 'food_rice', 'curtain', 'centerpiece', 'towel_rack', 'hair', 'clothes_dress', 'napkin', 'bathroom_cabinet', 'home_office', 'table', 'soap', 'shoe_rack']\nTask:Check appearance in mirror\nresponse:\n{\n    \"thoughts\": Walk to full-length mirror. Look in mirror. Make sure everything about your appearance is adequate. Adjust clothes as needed.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <mirror>', '[FIND] <mirror>', '[TURNTO] <mirror>', '[LOOKAT] <mirror>', '[FIND] <face>', '[TURNTO] <face>', '[LOOKAT] <face>', '[FIND] <hair>', '[TURNTO] <hair>', '[LOOKAT] <hair>', '[FIND] <clothes_dress>', '[TURNTO] <clothes_dress>', '[LOOKAT] <clothes_dress>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'stovefan', 'wall_clock', 'powersocket', 'bed', 'food_food', 'television', 'wooden_spoon', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'hairdryer', 'home_office', 'light', 'clothes_skirt', 'bathtub', 'bathroom_cabinet']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up your book and read it", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'cards', 'dresser', 'door', 'colander']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'detergent', 'toothbrush', 'orchid', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'toaster', 'ceilinglamp', 'window', 'teeth', 'couch', 'pillow', 'nightstand', 'wall_clock', 'wall', 'powersocket', 'cup', 'photoframe', 'toothbrush_holder', 'doorjamb', 'towel_rack', 'home_office', 'table', 'tooth_paste', 'coffee']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, grab the toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now wash the teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'door', 'table', 'light', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go in the living room. Then I find the remote, sit on the couch and turn on the television. Then I use the remote the change the TV channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'measuring_cup', 'chair', 'trashcan', 'toothbrush', 'cpuscreen', 'hanger', 'ceiling', 'character', 'keyboard', 'alcohol', 'mat', 'closetdrawer', 'form', 'dining_room', 'nightstand', 'oven', 'bench', 'bed', 'photoframe', 'food_food', 'coffe_maker', 'toothbrush_holder', 'tray', 'towel_rack', 'kitchen_counter', 'bathtub', 'sink', 'home_office', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'crayon', 'freezer', 'bathroom', 'food_carrot', 'drawing', 'knifeblock', 'floor', 'desk', 'ceilinglamp', 'stereo', 'window', 'couch', 'food_orange', 'pillow', 'form', 'dining_room', 'wall', 'bench', 'coffe_maker', 'television', 'rag', 'pencil', 'comb', 'table', 'tooth_paste', 'clothes_shirt', 'scrabble', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['water_glass', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'cupboard', 'sink', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Drink a glass of water in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'milk', 'drawing', 'bathroom_counter', 'floor', 'phone', 'character', 'bedroom', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'check', 'television', 'curtain', 'mouse', 'door', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Do homework\nresponse:\n{\n    \"thoughts\": Sit on the couch in the Living room and pick u some papers,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <check>', '[GRAB] <check>', '[DROP] <check>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'oven', 'cutting_board', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'dresser', 'food_steak', 'sink', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I sit at my computer desk and open the browser.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'chair', 'detergent', 'food_carrot', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'bookmark', 'toaster', 'balanceball', 'wall', 'powersocket', 'photoframe', 'coffee_filter', 'tray', 'doorjamb', 'cards', 'mouse', 'shaving_cream', 'shoes', 'dresser', 'cupboard', 'door', 'napkin', 'home_office', 'light', 'vase']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on light in an area,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'bathroom', 'toothbrush', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'phone', 'bedroom', 'computer', 'button', 'toaster', 'mat', 'stereo', 'foundation', 'window', 'couch', 'bench', 'oven', 'cup', 'check', 'television', 'tray', 'dresser', 'home_office', 'light', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room, look for radio, grab on button", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_ice_cream', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'longboard', 'floor', 'bookshelf', 'computer', 'faucet', 'balanceball', 'window', 'iron', 'food_peanut_butter', 'food_jam', 'wall', 'nightstand', 'bench', 'powersocket', 'basket_for_clothes', 'novel', 'check', 'coffee_filter', 'shaving_cream', 'pencil', 'maindoor', 'sink', 'hairdryer']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter the Home office, then walk to the phone, pickup the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'bench', 'bed', 'coffe_maker', 'television', 'food_donut', 'doorjamb', 'mouse', 'door', 'table', 'shoe_rack']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to living-room. Look at the desk where Computer stays. Sit in chair. Switch on computer. Work on computer. Switch off computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[TURNTO] <desk>", "[LOOKAT] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[WATCH] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'ceilingfan', 'drawing', 'needle', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'clothes_socks', 'couch', 'pillow', 'wall', 'powersocket', 'food_cheese', 'bed', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit down on the sofa. Pick up the remote control. Turn on the television with the remote control. Put down the remote control. Watch the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'bed', 'television', 'doorjamb', 'centerpiece', 'mouse', 'door', 'bathtub', 'sink', 'napkin', 'table', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Sit down on chair in front of computer, turn on computer, find mouse.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'detergent', 'trashcan', 'orchid', 'wallshelf', 'bathroom_counter', 'ceiling', 'keyboard', 'stovefan', 'tvstand', 'shower', 'window', 'mousepad', 'dining_room', 'wall', 'basket_for_clothes', 'television', 'centerpiece', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'clothes_dress', 'laundry_detergent', 'door', 'sink', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the washing machine. Open the washing machine. Find the basket for clothes. Find the clothes_dress. Grab the clothes_dress. Put the clothes_dress into the washing machine. Find the clothes_shirt. Grab the clothes_shirt. Put the clothes_shirt into the washing machine. Close the washing machine. Switch on the washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <basket_for_clothes>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'keyboard', 'envelope', 'stovefan', 'toaster', 'faucet', 'shower', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'bench', 'food_vegetable', 'bed', 'coffe_maker', 'check', 'water', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "walk to kicthen , find stove , find cooking pot , find vegetables , find water  ,put vegetables in cooking pot, pour water to cooking pot,put cooking pot on stove, switch on stove", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[FIND] <pot>", "[FIND] <food_vegetable>", "[FIND] <water>", "[GRAB] <water>", "[WASH] <food_vegetable>", "[POUR] <water> <pot>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'sheets', 'filing_cabinet', 'food_bread', 'toilet', 'ground_coffee', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'alcohol', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'plate', 'iron', 'mousepad', 'wall', 'oven', 'photoframe', 'light_bulb', 'coffe_maker', 'centerpiece', 'mouse', 'dresser', 'maindoor']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[TURNTO] <toilet>', '[LOOKAT] <toilet>', '[WALK] <toilet>', '[TURNTO] <toilet>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'food_cake', 'envelope', 'computer', 'ceilinglamp', 'window', 'mousepad', 'clothes_socks', 'closetdrawer', 'wall', 'toy', 'cup', 'coffe_maker', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I turn on computer. I use keyboard to type. I open browser. I search.", "actions": ["[WALK] <bedroom>", "[WALK] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'toaster', 'window', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'wall_clock', 'television', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'clothes_shirt']\nTask:Organize closet\nresponse:\n{\n    \"thoughts\": Organize the hanging clothes in my closet by color and type and get rid of clothes I don't wear.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[CLOSE] <dresser>', '[WALK] <basket_for_clothes>', '[PUTBACK] <clothes_shirt> <basket_for_clothes>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'plate', 'couch', 'pillow', 'wall', 'newspaper', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'bathtub', 'light', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "My cat looks like it wants attention, so I will go pet it on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'measuring_cup', 'toilet', 'freezer', 'bathroom', 'cpuscreen', 'character', 'desk', 'bedroom', 'table_cloth', 'mat', 'shower', 'stereo', 'microphone', 'clothes_hat', 'pillow', 'nightstand', 'wall_clock', 'oven', 'keys', 'light_bulb', 'check', 'razor', 'tray', 'doorjamb', 'dresser', 'clothes_dress', 'bag', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I find a radio. I turn on the radio. I listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'detergent', 'cat', 'drawing', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'dustpan', 'pillow', 'wall', 'wall_clock', 'television', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'cupboard', 'table', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I see my cat on the couch so I walk over, sit down and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'toilet', 'chair', 'bathroom', 'orchid', 'knifeblock', 'floor', 'phone', 'desk', 'keyboard', 'food_cake', 'foundation', 'floor_lamp', 'pillow', 'nightstand', 'powersocket', 'box', 'basket_for_clothes', 'razor', 'tray', 'doorjamb', 'cupboard', 'bathtub', 'spoon', 'home_office', 'table', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to living room, find lamp, switch on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'freezer', 'band-aids', 'cat', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'phone', 'folder', 'desk', 'tea_bag', 'keyboard', 'stovefan', 'toaster', 'shower', 'faucet', 'couch', 'bench', 'photoframe', 'coffe_maker', 'television', 'tray', 'vacuum_cleaner', 'hairbrush', 'dresser', 'door', 'bathtub', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat; pet the cat by touching it", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'band-aids', 'cat', 'food_carrot', 'knife', 'foundation', 'electric_shaver', 'mousepad', 'clothes_socks', 'pillow', 'dining_room', 'nightstand', 'oven', 'toy', 'check', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'chessboard', 'cupboard', 'conditioner', 'door', 'food_steak', 'clothes_skirt', 'duster']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Get a broom or duster with long handle. Dust blades of ceiling fan.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <duster>', '[FIND] <duster>', '[GRAB] <duster>', '[PUTOBJBACK] <duster>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'trashcan', 'drawing', 'floor', 'phone', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'shower', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'novel', 'coffe_maker', 'television', 'tray', 'curtain', 'cupboard', 'controller', 'door', 'bathtub', 'table', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to the couch in the living room, sit on ti and read the book", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'detergent', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bookshelf', 'mail', 'toaster', 'stovefan', 'faucet', 'closetdrawer', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'oven', 'photoframe', 'television', 'coin', 'drinking_glass', 'water', 'curtain', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'sink', 'dishwasher', 'coffee']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and approach the sink. I take a drinking glass and set it under the faucet. I turn the faucet until the water fills the glass, then turn it back. Finally I lift the glass out of the sink and drink the water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[SWITCHOFF] <faucet>', '[GRAB] <drinking_glass>', '[FIND] <water>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'desk', 'envelope', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'oven_mitts', 'curtain', 'controller', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I cross the room to the couch. I sit down and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'tablelamp', 'mop', 'toilet', 'walllamp', 'cat', 'drawing', 'floor', 'ceiling', 'character', 'bookshelf', 'food_cake', 'piano_bench', 'shower', 'faucet', 'pillow', 'wall', 'photoframe', 'bed', 'curtain', 'doorjamb', 'diary', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'soap', 'instrument_violin']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'mat', 'shower', 'window', 'food_kiwi', 'wall', 'bench', 'newspaper', 'food_food', 'coffe_maker', 'food_onion', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'home_office']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'freezer', 'walllamp', 'food_butter', 'dough', 'food_carrot', 'drawing', 'food_egg', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'food_chicken', 'food_cake', 'stovefan', 'mat', 'faucet', 'oil', 'wall', 'pasta', 'powersocket', 'oven', 'wall_clock', 'cutting_board', 'food_food', 'coffe_maker', 'oven_mitts', 'wooden_spoon', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'maindoor', 'sink', 'bathroom_cabinet', 'light']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and walk over the cabinent. I pull out a pan and walk to the stove. I put the pan on the stove and walk over to the fridge. I pull out out meat from the fridge and walk back over the stove. Then I put the meat in the pan.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[WALK] <oven>', '[PUTBACK] <pot> <oven>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[WALK] <oven>', '[PUTBACK] <food_chicken> <oven>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'filing_cabinet', 'nail_polish', 'toothbrush', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'food_food', 'toothbrush_holder', 'tray', 'dresser', 'cupboard', 'controller', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'light', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I go into the bathroom, go to the sink, get toothbrush and toothpaste. Put toothpaste on the toothbrush, brush teeth, turn on faucet, rinse brush, and return toothbrush.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'toilet', 'walllamp', 'cat', 'food_carrot', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'mop_bucket', 'stovefan', 'mat', 'stereo', 'couch', 'wall', 'powersocket', 'bed', 'novel', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'controller', 'door', 'food_steak', 'spoon', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.,\n    \"actions\": ['[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'floor', 'ceiling', 'food_chicken', 'stovefan', 'mat', 'toaster', 'shower', 'faucet', 'window', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'scissors', 'microwave', 'bathroom', 'nail_polish', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'floor', 'bookshelf', 'mat', 'faucet', 'window', 'electric_shaver', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'photoframe', 'newspaper', 'keys', 'television', 'curtain', 'doorjamb', 'diary', 'sink', 'spoon', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'drawing', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'food_food', 'coffe_maker', 'television', 'tray', 'curtain', 'mouse', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'sink', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I sit in the sofa and I relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'toilet', 'chair', 'walllamp', 'ceilingfan', 'bathroom_counter', 'floor', 'phone', 'character', 'mail', 'computer', 'toaster', 'foundation', 'clothes_socks', 'food_peanut_butter', 'dining_room', 'wall', 'oven', 'powersocket', 'tape', 'cup', 'coffe_maker', 'video_game_controller', 'television', 'tray', 'curtain', 'cd', 'maindoor', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'jelly', 'dog', 'walllamp', 'bathroom', 'cpuscreen', 'hanger', 'bedroom', 'keyboard', 'toaster', 'ceilinglamp', 'stereo', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'oven', 'photoframe', 'coffe_maker', 'television', 'curtain', 'mouse', 'cupboard', 'bathtub', 'home_office', 'light', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the bedroom. I walk to stereo. I turn on stereo.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'mousepad', 'couch', 'towel', 'wall', 'bed', 'photoframe', 'curtain', 'after_shave', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'light', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap and hot water,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'shoe_rack', 'tablelamp', 'microwave', 'toilet', 'walllamp', 'bathroom', 'orchid', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'shower', 'mat', 'window', 'dish_soap', 'couch', 'pillow', 'wall', 'basket_for_clothes', 'bed', 'towel_rack', 'light', 'dishwasher', 'creditcard']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Wash a plate using the dishwasher in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'crayon', 'cat', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'food_cake', 'computer', 'brush', 'electric_shaver', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'toy', 'food_food', 'towel_rack', 'mouse', 'dresser', 'pajamas', 'sink', 'napkin', 'light', 'colander', 'tooth_paste']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTBACK] <toy> <dresser>', '[CLOSE] <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'bathroom', 'food_butter', 'milk', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'bookshelf', 'computer', 'ceilinglamp', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'clothes_hat', 'wall', 'coffe_maker', 'television', 'tray', 'curtain', 'dresser', 'bathtub', 'bathroom_cabinet', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the entrance hall. Find the telephone. Grab the telephone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'food_butter', 'knife', 'wallshelf', 'bathroom_counter', 'phone', 'character', 'bedroom', 'keyboard', 'computer', 'faucet', 'electric_shaver', 'mousepad', 'wall', 'powersocket', 'oven', 'shelf', 'cup', 'bed', 'food_rice', 'tray', 'mouse', 'headset', 'comb', 'dresser', 'conditioner', 'home_office', 'table']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": go to the bathroom and brush your teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>']\n}\n", "input": "\nScene_Objects:['food_salt', 'freezer', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'television', 'ice', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'headset', 'cupboard', 'sink', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the sofa, and lie down on the sofa.", "actions": ["[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'detergent', 'toothbrush', 'drawing', 'cpuscreen', 'food_egg', 'phone', 'character', 'computer', 'piano_bench', 'shoe-shine_kit', 'faucet', 'food_noodles', 'mousepad', 'clothes_hat', 'towel', 'broom', 'cutting_board', 'shelf', 'keys', 'novel', 'food_food', 'check', 'wooden_spoon', 'towel_rack', 'sink', 'table', 'light', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'freezer', 'walllamp', 'cpuscreen', 'wallshelf', 'ceiling', 'computer', 'shower', 'window', 'mousepad', 'couch', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'bench', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'razor', 'mouse', 'cupboard', 'laundry_detergent', 'door', 'home_office', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "\"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\"", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[TURNTO] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'freezer', 'walllamp', 'ground_coffee', 'bathroom', 'ceilingfan', 'clothes_underwear', 'bathroom_counter', 'character', 'thread', 'piano_bench', 'toaster', 'deck_of_cards', 'homework', 'couch', 'dustpan', 'powersocket', 'newspaper', 'clothes_gloves', 'novel', 'coffe_maker', 'doorjamb', 'chessboard', 'cd', 'napkin', 'home_office', 'instrument_guitar', 'clothes_skirt', 'light']\nTask:Close door\nresponse:\n{\n    \"thoughts\": Push the door shut until it latches.,\n    \"actions\": ['[RUN] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'jelly', 'freezer', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'computer', 'food_bacon', 'plate', 'faucet', 'shower', 'window', 'dish_soap', 'mousepad', 'couch', 'wall', 'wall_clock', 'powersocket', 'bench', 'shoes', 'hairbrush', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Make some dish water, grab a dish rag, put dishes in dish water and wash.", "actions": ["[RUN] <sink>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[PUTBACK] <dish_soap> <plate>", "[RINSE] <dish_soap>", "[FIND] <plate>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'mop', 'bathroom', 'knife', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'desk', 'food_cake', 'window', 'electric_shaver', 'microphone', 'food_jam', 'dining_room', 'wall', 'photoframe', 'food_food', 'television', 'curtain', 'mouse', 'kitchen_counter', 'cd', 'chessboard', 'dresser', 'hairbrush', 'clothes_scarf', 'bathroom_cabinet', 'dishwasher', 'instrument_violin']\nTask:Put on coat\nresponse:\n{\n    \"thoughts\": I remove my scarf from a coat peg in the entrance hall.  I wrap the scarf around my neck,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <hanger>', '[TURNTO] <hanger>', '[LOOKAT] <hanger>', '[FIND] <clothes_scarf>', '[GRAB] <clothes_scarf>', '[PUTON] <clothes_scarf>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'walllamp', 'cat', 'trashcan', 'milk', 'hanger', 'bathroom_counter', 'desk', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'dining_room', 'nightstand', 'wall', 'oven', 'cup', 'bed', 'coffe_maker', 'tray', 'curtain', 'chessboard', 'shoes', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting a cat may sound simple.Letting the Cat Come to You.Extend a hand or finger and allow the cat a chance to touch her nose to you. When a cat bumps her head into your hand, it's a signal she wants attention.Focusing on Areas with Scent Glands. Feed with food.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <milk>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[TURNTO] <cat>", "[POINTAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'microwave', 'toilet', 'toothbrush', 'bills', 'orchid', 'character', 'bedroom', 'stamp', 'stovefan', 'laptop', 'ceilinglamp', 'shower', 'mat', 'faucet', 'towel', 'pillow', 'dining_room', 'oven', 'coffe_maker', 'television', 'toothbrush_holder', 'comb', 'cupboard', 'door', 'sink', 'table', 'soap', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find toothbrush.  Find toothpaste.  Put toothpaste on toothbrush.  Brush teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>', '[FIND] <teeth>', '[SCRUB] <teeth>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_salt', 'microwave', 'freezer', 'orchid', 'cpuscreen', 'ceiling', 'character', 'bedroom', 'desk', 'food_bacon', 'bookmark', 'toaster', 'stovefan', 'faucet', 'shower', 'mousepad', 'wall', 'wall_clock', 'oven', 'bench', 'coffe_maker', 'television', 'water', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen, take a glass from the cupboard, fill with water and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>", "[RINSE] <water_glass>", "[PUTBACK] <water_glass> <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'walllamp', 'detergent', 'cat', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'remote_control', 'mail', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'couch', 'wall', 'bench', 'clothes_gloves', 'coffe_maker', 'television', 'curtain', 'coffee_pot', 'home_office', 'table', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'bathroom', 'cat', 'dough', 'knife', 'bathroom_counter', 'floor', 'keyboard', 'computer', 'brush', 'laptop', 'plate', 'faucet', 'deck_of_cards', 'window', 'closetdrawer', 'couch', 'dining_room', 'nightstand', 'wall', 'oven', 'bench', 'television', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathtub']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Every time this animal is seen, you must approach it and show compassion for it by petting it in a gentle manner.", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <bedroom>", "[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[FIND] <brush>", "[GRAB] <brush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'blender', 'microwave', 'toilet', 'detergent', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'character', 'keyboard', 'computer', 'brush', 'mat', 'laptop', 'faucet', 'ceilinglamp', 'window', 'dustpan', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'towel_rack', 'mouse', 'kitchen_counter', 'pajamas', 'dishwasher']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go to bedroom, put pajamas on, lie in bed, sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <pajamas>', '[FIND] <pajamas>', '[GRAB] <pajamas>', '[PUTON] <pajamas>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'window', 'mousepad', 'wall', 'nightstand', 'powersocket', 'bench', 'photoframe', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'cd', 'sink', 'bathtub', 'table', 'dirt']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to my office and sit at my desk. Turn on the computer and type my email at the keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'knife', 'knifeblock', 'orchid', 'drawing', 'floor', 'phone', 'character', 'bedroom', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'faucet', 'shower', 'oil', 'dining_room', 'powersocket', 'newspaper', 'ice', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'door', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I pick up the remote. I press button on remote.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'toilet', 'freezer', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'ceiling', 'phone', 'character', 'keyboard', 'toaster', 'mat', 'floor_lamp', 'oil', 'pillow', 'dining_room', 'nightstand', 'wall_clock', 'bench', 'novel', 'television', 'curtain', 'doorjamb', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Turn a light on and read a book in the bedroom.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'bathroom', 'trashcan', 'food_butter', 'drawing', 'floor', 'ceiling', 'phone', 'mat', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'food_food', 'check', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'dresser', 'door', 'sink', 'spoon', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'crayon', 'toilet', 'bathroom', 'nail_polish', 'trashcan', 'food_carrot', 'hanger', 'floor', 'computer', 'toaster', 'window', 'food_jam', 'nightstand', 'oven', 'bench', 'tape', 'photoframe', 'food_food', 'coffe_maker', 'television', 'toilet_paper', 'doorjamb', 'dresser', 'pajamas', 'cupboard', 'door', 'sink', 'home_office']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the living room. I find the stereo and turn it on, and I listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'microwave', 'chair', 'detergent', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'purse', 'computer', 'toaster', 'mat', 'tvstand', 'clothes_socks', 'food_peanut_butter', 'microphone', 'food_jam', 'pillow', 'wall', 'powersocket', 'after_shave', 'doorjamb', 'cd', 'pajamas', 'laundry_detergent', 'door', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Coming home turning on lights.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'trashcan', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'oil', 'couch', 'wall', 'wall_clock', 'powersocket', 'oven', 'food_food', 'television', 'doorjamb', 'mouse', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "turn computer on, open application to work with", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'food_bread', 'scissors', 'microwave', 'freezer', 'milk', 'knifeblock', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'mail', 'food_bacon', 'ceilinglamp', 'shower', 'faucet', 'bowl', 'food_kiwi', 'wall', 'oven', 'cup', 'bed', 'coffe_maker', 'video_game_controller', 'diary', 'dishrack', 'controller', 'conditioner', 'home_office']\nTask:Get glass of milk\nresponse:\n{\n    \"thoughts\": Get an empty glass. Take milk from refrigerator and open it. Pour milk into glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[POUR] <milk> <cup>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'piano_bench', 'stovefan', 'mat', 'tvstand', 'shower', 'stereo', 'mousepad', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'bench', 'photoframe', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'sink', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Grab groceries and walk to the fridge. Open fridge and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'laptop', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dishrack', 'door', 'light', 'shoe_rack']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office, walk to desk, sit in chair, turn on laptop, grab mouse, open a web page to look at it, move mouse, keep looking at laptop screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[TOUCH] <mouse>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'bookshelf', 'mail', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'oven', 'toy', 'mouthwash', 'television', 'wine', 'tray', 'curtain', 'cd', 'dresser', 'bathtub', 'sink', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room,find the sofa,sat down in the sofa,take some relaxation.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'photoframe', 'novel', 'television', 'mouse', 'kitchen_counter', 'door', 'sink', 'clothes_skirt', 'clothes_shirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to home office, find the book. Sit in the chair and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'shower', 'window', 'tvstand', 'pillow', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'coffe_maker', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'bathroom', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bookshelf', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'couch', 'dining_room', 'food_kiwi', 'wall', 'toilet_paper', 'razor', 'television', 'curtain', 'diary', 'towel_rack', 'door', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch and use the remote control to turn on the TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'orchid', 'hanger', 'floor', 'phone', 'ceiling', 'character', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'window', 'closetdrawer', 'powersocket', 'oven', 'cup', 'curtain', 'doorjamb', 'mouse', 'comb', 'cupboard', 'clothes_dress', 'door', 'bathtub', 'home_office', 'table', 'dishwasher', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'beer', 'crayon', 'walllamp', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'bookshelf', 'mat', 'iron', 'electric_shaver', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'toy', 'cup', 'light_bulb', 'novel', 'food_food', 'check', 'toothbrush_holder', 'television', 'curtain', 'doorjamb', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the book case, I select a book, I go back and sit down, then I read.,\n    \"actions\": ['[WALK] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'cpuscreen', 'hanger', 'bathroom_counter', 'phone', 'bedroom', 'desk', 'bookshelf', 'toaster', 'ceilinglamp', 'mat', 'faucet', 'window', 'mousepad', 'closetdrawer', 'pillow', 'dining_room', 'broom', 'nightstand', 'oven', 'bench', 'coffe_maker', 'check', 'towel_rack', 'mouse', 'cupboard', 'light', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I find a radio. I turn on the radio. I listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_bread', 'food_ice_cream', 'crayon', 'microwave', 'chair', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'faucet', 'tvstand', 'window', 'wall', 'bench', 'powersocket', 'clothes_gloves', 'water', 'tray', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'light', 'instrument_violin']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen, hold drinking glass, lift glass, pour water in drinking glass,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[FIND] <water>', '[GRAB] <water>', '[WALK] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'freezer', 'bathroom', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'computer', 'ceilinglamp', 'mat', 'faucet', 'couch', 'clothes_pants', 'wall', 'wall_clock', 'bench', 'powersocket', 'nightstand', 'photoframe', 'novel', 'tray', 'curtain', 'doorjamb', 'sink', 'napkin', 'bathroom_cabinet', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Going to the sink and scrubbing hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'orchid', 'wallshelf', 'clothes_underwear', 'cpuscreen', 'phone', 'character', 'bedroom', 'keyboard', 'toaster', 'mat', 'plate', 'picture', 'foundation', 'dish_soap', 'clothes_socks', 'pillow', 'powersocket', 'cutting_board', 'basket_for_clothes', 'bed', 'television', 'curtain', 'diary', 'cupboard', 'sink', 'spoon', 'home_office']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Pick up dish.  Put in soapy water.  Wash dish.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dish_soap>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <dish_soap> <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WIPE] <plate>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'wallshelf', 'phone', 'desk', 'toaster', 'stovefan', 'mat', 'shower', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'powersocket', 'bench', 'bed', 'photoframe', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'bathtub', 'sink', 'bathroom_cabinet', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the cat. Pet the cat with your hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'floor', 'character', 'keyboard', 'bookshelf', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'wall_clock', 'toilet_paper', 'mouthwash', 'television', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, find sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['water_glass', 'microwave', 'freezer', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'button', 'toaster', 'stovefan', 'mat', 'faucet', 'shower', 'couch', 'wall', 'wall_clock', 'oven', 'keys', 'coffe_maker', 'water', 'tray', 'curtain', 'coffee_pot', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'chair', 'walllamp', 'orchid', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'desk', 'bedroom', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'iron', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'bed', 'television', 'doorjamb', 'mouse', 'door', 'home_office', 'light', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": enter office, find computer, turn it on,,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TOUCH] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'blender', 'microwave', 'toilet', 'freezer', 'detergent', 'cat', 'food_carrot', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'ceilinglamp', 'faucet', 'oil', 'dining_room', 'nightstand', 'wall_clock', 'oven', 'bed', 'television', 'coffee_filter', 'tray', 'mouse', 'dresser', 'hair', 'light', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into my room and turn the radio on so that I can listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'detergent', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'laptop', 'shower', 'window', 'oil', 'clothes_socks', 'couch', 'food_orange', 'pillow', 'wall', 'novel', 'coffe_maker', 'television', 'food_rice', 'rag', 'doorjamb', 'cupboard', 'door', 'sink', 'napkin', 'table', 'light']\nTask:Relax\nresponse:\n{\n    \"thoughts\": Lie down on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'measuring_cup', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'bench', 'cup', 'photoframe', 'novel', 'check', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'bathroom_cabinet', 'light', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I go to the book case, I select a book, I go back and sit down, then I read.", "actions": ["[WALK] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'chair', 'freezer', 'walllamp', 'food_carrot', 'drawing', 'knife', 'hanger', 'floor', 'phone', 'ceiling', 'keyboard', 'computer', 'brush', 'deck_of_cards', 'food_jam', 'bowl', 'dining_room', 'wall', 'bench', 'razor', 'doorjamb', 'mouse', 'dishrack', 'dresser', 'cupboard', 'door', 'sink', 'table', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Pick up phone,\n    \"actions\": ['[WALK] <table>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'toaster', 'mat', 'tvstand', 'mousepad', 'pillow', 'dining_room', 'wall', 'coffee_table', 'novel', 'television', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'cupboard', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I walk to the living room and pick up a book from the coffee table. I find my favorite chair.  I sit in the armchair and open the book.  I read.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[WALK] <coffee_table>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'microwave', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'button', 'food_bacon', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'coffee_table', 'novel', 'mouthwash', 'television', 'doorjamb', 'diary', 'clothes_scarf', 'sink']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit down in recliner.  Pick up novel off of coffee table.  Open novel to last read page.  Begin reading,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <bookmark>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'freezer', 'walllamp', 'ceiling', 'bedroom', 'keyboard', 'toaster', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'clothes_pants', 'food_orange', 'nightstand', 'bench', 'photoframe', 'bed', 'food_food', 'tray', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'bathtub', 'bathroom_cabinet', 'table', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk into laundry room. Open washing machine door. Put dirty clothes into washing machine. Measure laundry detergent and place into washing machine. Turn dial to appropriate wash setting. Power on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[CLOSE] <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[OPEN] <washing_machine>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'freezer', 'knifeblock', 'drawing', 'wallshelf', 'ceiling', 'phone', 'desk', 'keyboard', 'face', 'toaster', 'laptop', 'tvstand', 'shower', 'stereo', 'homework', 'oil', 'towel', 'oven', 'basket_for_clothes', 'tray', 'towel_rack', 'hair', 'cupboard', 'door', 'bathtub', 'clothes_shirt']\nTask:Rain welcome\nresponse:\n{\n    \"thoughts\": The robot would greet me and then hand me a towel.  It would make an appropriate quip about the weather while I dried off.  It would receive the wet towel when I was done.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <towel>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <face>', '[WIPE] <face>', '[FIND] <hair>', '[WIPE] <hair>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'trashcan', 'toothbrush', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'alcohol', 'stovefan', 'faucet', 'shower', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'bench', 'photoframe', 'towel_rack', 'mouse', 'cupboard', 'conditioner', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'home_office', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Upon entering the restroom, walk to the sink. Take a toothbrush and apply enough toothpaste so that 1/4 of the brush is covered (this is the appropriate amount). Proceed to \"brush\" your teeth using an up and down motion with the brush in your mouth. Brush for 2 minutes covering all areas of the teeth. Rinse with water and clean toothbrush with water too.", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[SCRUB] <tooth_paste>", "[FIND] <toothbrush>", "[WASH] <toothbrush>", "[FIND] <sink>", "[RINSE] <sink>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_bread', 'toilet', 'chair', 'freezer', 'bathroom', 'food_carrot', 'hanger', 'keyboard', 'envelope', 'mail', 'tvstand', 'clothes_pants', 'bowl', 'dining_room', 'oven', 'powersocket', 'bed', 'clothes_gloves', 'check', 'wine', 'curtain', 'food_onion', 'kettle', 'dresser', 'vase', 'home_office', 'light', 'clothes_skirt', 'dishwasher']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I hear the phone ringing in the living room. I walk to the living room and go to the phone. I pick it up and answer the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>', '[SWITCHON] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'chair', 'walllamp', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'laptop', 'shower', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'oven', 'cutting_board', 'coffe_maker', 'television', 'tray', 'doorjamb', 'mouse', 'hair', 'door', 'sink', 'home_office', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I walk to the Home Office then find the computer and sit in the chair. I switch on the computer and start typing the email. I switch the computer off.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'detergent', 'bills', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'mat', 'laptop', 'shower', 'faucet', 'ceilinglamp', 'mousepad', 'microphone', 'form', 'pillow', 'broom', 'wall', 'newspaper', 'after_shave', 'doorjamb', 'cd', 'comb', 'controller', 'door', 'bathroom_cabinet', 'light', 'table', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.,\n    \"actions\": ['[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'bedroom', 'computer', 'toaster', 'mat', 'oil', 'couch', 'dining_room', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'food_rice', 'tray', 'curtain', 'mouse', 'dresser', 'bathtub', 'home_office', 'table', 'bathroom_cabinet', 'coffee', 'shoe_rack']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I pick up my phone. I touch it to navigate to the web browser. I can swipe and tap the screen to visit different websites. Then I read or watch the content within those websites.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[TOUCH] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'tea_bag', 'ceilinglamp', 'tvstand', 'shower', 'window', 'homework', 'couch', 'dustpan', 'food_orange', 'dining_room', 'wall', 'nightstand', 'photoframe', 'clothes_gloves', 'coffe_maker', 'check', 'television', 'mouse', 'pencil', 'table', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, turn to the sofa. Sit on the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'nail_polish', 'knife', 'cpuscreen', 'bathroom_counter', 'table_cloth', 'stovefan', 'tvstand', 'stereo', 'faucet', 'window', 'mousepad', 'dining_room', 'broom', 'nightstand', 'wall_clock', 'wall', 'oven', 'bed', 'newspaper', 'food_food', 'video_game_controller', 'wine', 'after_shave', 'curtain', 'towel_rack', 'bathtub', 'light', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find radio, turn it on", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'jelly', 'bathroom', 'detergent', 'dough', 'orchid', 'floor', 'bedroom', 'mat', 'dish_soap', 'mousepad', 'dining_room', 'wall', 'nightstand', 'oven', 'food_cheese', 'photoframe', 'coffe_maker', 'food_food', 'ice', 'tray', 'kitchen_counter', 'headset', 'dishrack', 'dresser', 'cupboard', 'door', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the dishwasher and put dish soap. I close the dishwasher and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'dustpan', 'pillow', 'wall', 'wall_clock', 'powersocket', 'television', 'tray', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Browsing the internet for information.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'detergent', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'bathroom_counter', 'phone', 'folder', 'man', 'keyboard', 'mop_bucket', 'tvstand', 'faucet', 'picture', 'plate', 'mousepad', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'check', 'food_onion', 'clothes_scarf', 'spoon', 'light', 'dishwasher']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": When someone knocks I n the door walk through the hallway and greet them.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'freezer', 'orchid', 'drawing', 'wallshelf', 'floor', 'phone', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'shower', 'window', 'mousepad', 'pillow', 'nightstand', 'wall', 'oven', 'wall_clock', 'bed', 'novel', 'curtain', 'towel_rack', 'cupboard', 'door', 'sink', 'light', 'laser_pointer']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the bedroom. Grab the book out of the nightstand. Lay down in bed. Open the book. Read the book.", "actions": ["[WALK] <bedroom>", "[WALK] <nightstand>", "[FIND] <nightstand>", "[OPEN] <nightstand>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[TURNTO] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'thread', 'computer', 'stovefan', 'mat', 'tvstand', 'toaster', 'couch', 'pillow', 'wall', 'photoframe', 'check', 'television', 'doorjamb', 'laundry_detergent', 'conditioner', 'sink', 'door', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Do homework\nresponse:\n{\n    \"thoughts\": Sit on the couch in the Living room and pick u some papers,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <check>', '[GRAB] <check>', '[DROP] <check>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'mop', 'chair', 'bathroom', 'drawing', 'wallshelf', 'character', 'bedroom', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'shower', 'faucet', 'floor_lamp', 'mousepad', 'closetdrawer', 'nightstand', 'wall', 'oven', 'bench', 'photoframe', 'bed', 'coffe_maker', 'food_food', 'check', 'conditioner', 'light', 'table', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room, find lamp, switch on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'jelly', 'freezer', 'walllamp', 'knifeblock', 'drawing', 'floor', 'ceiling', 'food_cake', 'stovefan', 'toaster', 'plate', 'mat', 'shower', 'food_kiwi', 'wall', 'nightstand', 'oven', 'powersocket', 'photoframe', 'food_food', 'tray', 'food_onion', 'doorjamb', 'kitchen_counter', 'sauce_pan', 'cupboard', 'door', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <spoon>', '[GRAB] <spoon>', '[FIND] <sauce_pan>', '[PUTBACK] <spoon> <sauce_pan>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <plate>', '[PUTBACK] <pot> <plate>']\n}\n", "input": "\nScene_Objects:['fork', 'toilet', 'freezer', 'detergent', 'drawing', 'knife', 'wallshelf', 'dvd_player', 'character', 'bookshelf', 'plate', 'shower', 'tvstand', 'dish_soap', 'closetdrawer', 'pillow', 'food_kiwi', 'nightstand', 'powersocket', 'cutting_board', 'photoframe', 'wooden_spoon', 'dishrack', 'dresser', 'cupboard', 'door', 'sink', 'spoon', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <dishrack>", "[PUTBACK] <fork> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishrack>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishrack>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'floor', 'ceiling', 'keyboard', 'stovefan', 'electric_shaver', 'clothes_socks', 'form', 'wall', 'nightstand', 'oven', 'cutting_board', 'coffe_maker', 'food_food', 'oven_mitts', 'food_donut', 'tray', 'doorjamb', 'kettle', 'mouse', 'kitchen_counter', 'laundry_detergent', 'home_office', 'table', 'bathroom_cabinet', 'instrument_guitar']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": go to the kitchen and put the groceries in the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_ice_cream', 'microwave', 'toilet', 'freezer', 'bathroom', 'trashcan', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'toaster', 'mat', 'shower', 'faucet', 'wall', 'nightstand', 'oven', 'food_food', 'video_game_controller', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk Into Kitchen. Open Refrigerator. Grab Food. Walk to Microwave. Insert food into microwave. Cook until warm.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[CLOSE] <freezer>", "[WALK] <microwave>", "[SWITCHON] <microwave>", "[SWITCHOFF] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'television', 'tray', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch youtube\nresponse:\n{\n    \"thoughts\": Go to youtube.Com watch videos,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'jelly', 'walllamp', 'orchid', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'envelope', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'wall', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'bathroom_cabinet', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go into the office and sit down at the desk in front of the computer. Turn on the computer and look at it while clicking with the mouse and typing with the keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'food_bacon', 'toaster', 'tvstand', 'window', 'oil', 'couch', 'wall', 'nightstand', 'newspaper', 'coffe_maker', 'television', 'coffee_filter', 'water', 'food_donut', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'dirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'thread', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'mat', 'foundation', 'mousepad', 'wall', 'powersocket', 'television', 'doorjamb', 'kettle', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I go to my home office, sit on the chair and turn on the computer. I open my email account to look for all the information I need.I pull up my online task on the computer to look at what the assigned task is.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'chair', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'ceiling', 'desk', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'homework', 'mousepad', 'closetdrawer', 'pillow', 'dining_room', 'wall_clock', 'oven', 'tape', 'photoframe', 'dresser', 'cupboard', 'bathtub', 'light', 'dirt']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I reach out arm. I put hand on radio. I turn on radio. I pick a station.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[TOUCH] <stereo>', '[TURNTO] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'tea_bag', 'food_bacon', 'mop_bucket', 'tvstand', 'electric_shaver', 'dining_room', 'wall', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'light', 'table', 'home_office', 'colander']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "push any ligth button to turn it on, for example in the living room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'microwave', 'walllamp', 'toothbrush', 'drawing', 'knifeblock', 'wallshelf', 'keyboard', 'food_cake', 'ceilinglamp', 'tvstand', 'shower', 'stereo', 'oil', 'tooth_paste', 'couch', 'closetdrawer', 'dining_room', 'toy', 'basket_for_clothes', 'light_bulb', 'check', 'centerpiece', 'kettle', 'mouse', 'kitchen_counter', 'shoes', 'light', 'board_game']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": EARLY MORNING AND BEFORE GOING TO BED,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['jelly', 'freezer', 'walllamp', 'bathroom', 'bills', 'needle', 'drawing', 'character', 'table_cloth', 'toaster', 'picture', 'floor_lamp', 'window', 'mousepad', 'microphone', 'food_kiwi', 'wall', 'wall_clock', 'newspaper', 'bed', 'television', 'tray', 'cupboard', 'controller', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TOUCH] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'computer', 'stovefan', 'tvstand', 'faucet', 'shower', 'pillow', 'wall', 'cup', 'bed', 'coffe_maker', 'television', 'tray', 'lighter', 'dresser', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Walk to the sink, turn on the faucet, put soap on hands, rub hands together under the water, rinse off hands, turn off faucet, dry hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[WALK] <sink>', '[FIND] <soap>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'bed', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'pencil', 'bathtub', 'home_office', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "For browsing internet walk to home office. Find Computer. Walk to Computer. Find Chair. Sit in a chair and switch on computer. Connect with internet and browse the net.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[WALK] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'bookmark', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'mousepad', 'dining_room', 'wall', 'nightstand', 'oven', 'cup', 'bed', 'television', 'coffee_filter', 'mouse', 'door']\nTask:Answer emails\nresponse:\n{\n    \"thoughts\": Open email application.  Open new emails.  Respond accordingly,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <mail>', '[GRAB] <mail>', '[READ] <mail>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <laptop>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'freezer', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'stovefan', 'toaster', 'tvstand', 'faucet', 'mousepad', 'towel', 'wall', 'wall_clock', 'bench', 'powersocket', 'nightstand', 'bed', 'toilet_paper', 'television', 'wine', 'curtain', 'doorjamb', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'clothes_skirt', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. Walk to the sink. Turn on tap. Get soap. Wash and rinse hands. Turn off tap. Dry hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[LOOKAT] <sink>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[RINSE] <soap>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'toilet', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'ceilinglamp', 'tvstand', 'stereo', 'faucet', 'mousepad', 'dining_room', 'wall', 'powersocket', 'food_food', 'food_donut', 'towel_rack', 'mouse', 'shoes', 'dishrack', 'sink', 'spoon', 'light', 'coffee']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to bathroom, find toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'dining_room', 'wall', 'oven', 'powersocket', 'food_food', 'coffe_maker', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'bathroom_cabinet']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "go to the kitchen and put the groceries in the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'love_seat', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'clothes_scarf']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit and watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'microwave', 'toilet', 'jelly', 'freezer', 'cpuscreen', 'floor', 'phone', 'ceiling', 'keyboard', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'foundation', 'clothes_socks', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'photoframe', 'tray', 'curtain', 'kitchen_counter', 'sink', 'clothes_scarf']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "walk to bathroom , run to toilet", "actions": ["[WALK] <bathroom>", "[RUN] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'chair', 'walllamp', 'drawing', 'needle', 'wallshelf', 'bills', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'form', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'check', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'tooth_paste']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I enter the living room and sit on the couch. I grab the remote and turn on the TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'water_glass', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'stovefan', 'mat', 'shower', 'closetdrawer', 'couch', 'wall', 'powersocket', 'oven', 'food_food', 'check', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen ,find fridge ,open fridge ,find water  glass , drink from water glass", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'detergent', 'orchid', 'clothes_underwear', 'floor', 'character', 'button', 'stovefan', 'ceilinglamp', 'toaster', 'homework', 'mousepad', 'couch', 'dustpan', 'microphone', 'form', 'towel', 'powersocket', 'basket_for_clothes', 'bed', 'light_bulb', 'oven_mitts', 'tray', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'colander', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to laundry room. Find laundry basket. Grab laundry basket. Find washing machine. Walk to washing machine. Open washing machine. Find detergent. Grab detergent. Put in detergent. Put in clothes. Close washing machine. Turn on machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[PUTOBJBACK] <laundry_detergent>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'character', 'bedroom', 'keyboard', 'table_cloth', 'toaster', 'stovefan', 'faucet', 'shower', 'window', 'mousepad', 'closetdrawer', 'wall', 'oven', 'lighter', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'home_office']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Travel through the house and locate the bathroom, and proceed to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'measuring_cup', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'bench', 'cup', 'photoframe', 'novel', 'check', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'bathroom_cabinet', 'light', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the book case, I select a book, I go back and sit down, then I read.,\n    \"actions\": ['[WALK] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'freezer', 'walllamp', 'drawing', 'orchid', 'needle', 'floor', 'phone', 'ceiling', 'character', 'keyboard', 'food_cake', 'computer', 'stovefan', 'mat', 'plate', 'window', 'closetdrawer', 'wall', 'powersocket', 'tape', 'television', 'tray', 'curtain', 'towel_rack', 'sink', 'spoon', 'bathtub', 'table', 'bathroom_cabinet']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Place your plate on the placemat. The forks will be placed on the left side of the plate and the knife and spoon on the right and place the spoon to the right of the knife. If you want to set out food, place them above the plate.", "actions": ["[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[DROP] <plate>", "[FIND] <fork>", "[GRAB] <fork>", "[DROP] <fork>", "[FIND] <spoon>", "[GRAB] <spoon>", "[DROP] <spoon>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Draft home\nresponse:\n{\n    \"thoughts\": Using mathematics and CAD software design a home,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'clothes_jacket', 'mat', 'ceilinglamp', 'shower', 'window', 'closetdrawer', 'couch', 'dining_room', 'wall', 'powersocket', 'food_food', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub', 'board_game']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Open fridge. Put groceries in fridge.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I sit at my desk and work on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'toilet', 'chair', 'freezer', 'drawing', 'floor', 'ceiling', 'keyboard', 'computer', 'stovefan', 'toaster', 'tvstand', 'faucet', 'shower', 'window', 'wall', 'oven', 'powersocket', 'cup', 'coffe_maker', 'coffee_filter', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'light', 'dirt', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'cat', 'floor', 'character', 'stamp', 'table_cloth', 'piano_bench', 'mat', 'window', 'couch', 'food_jam', 'pillow', 'wall', 'nightstand', 'bench', 'toy', 'cup', 'mouthwash', 'television', 'coffee_filter', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'clothes_dress', 'door', 'food_steak', 'bathtub', 'bag', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I see my cat on the couch so I walk over, sit down and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['ground_coffee', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'toaster', 'mat', 'faucet', 'electric_shaver', 'towel', 'pillow', 'wall', 'wall_clock', 'bench', 'bed', 'coffe_maker', 'curtain', 'doorjamb', 'dresser', 'sink', 'bag', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to bathroom, turn on faucet, wet hands, apply soap, rub hands, rinse hands, turn off faucet, dry hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'toilet', 'freezer', 'dough', 'drawing', 'hanger', 'floor', 'ceiling', 'mat', 'shoe-shine_kit', 'shower', 'clothes_hat', 'bowl', 'nightstand', 'wall', 'powersocket', 'wall_clock', 'photoframe', 'light_bulb', 'food_food', 'coffe_maker', 'razor', 'coffee_filter', 'doorjamb', 'hairbrush', 'door', 'home_office', 'light', 'soap']\nTask:Put away groceries\nresponse:\n{\n    \"thoughts\": Place groceries in refrigerator from shopping at the grocery store.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'brush', 'ceilinglamp', 'tvstand', 'plate', 'mousepad', 'closetdrawer', 'dustpan', 'pillow', 'wall', 'powersocket', 'television', 'tray', 'doorjamb', 'centerpiece', 'mouse', 'shoes', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "walk to living room, look at computer, find chair, sit in chair, grab keyboard, grab mouse, type using keyboard and mouse", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[WALK] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'crayon', 'walllamp', 'ceilingfan', 'dough', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'mat', 'stovefan', 'wall', 'powersocket', 'oven', 'bench', 'cutting_board', 'food_rice', 'doorjamb', 'cd', 'door', 'light', 'vase', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I enter the bedroom and find the light. Then I turn it on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'bookshelf', 'stovefan', 'mat', 'shower', 'window', 'closetdrawer', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'wall_clock', 'photoframe', 'food_food', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'home_office']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Put groceries into the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'chair', 'freezer', 'walllamp', 'detergent', 'orchid', 'phone', 'bookshelf', 'computer', 'toaster', 'tvstand', 'stereo', 'closetdrawer', 'food_peanut_butter', 'clothes_pants', 'form', 'wall', 'oven', 'cup', 'bed', 'mouthwash', 'rag', 'towel_rack', 'kitchen_counter', 'pajamas', 'food_steak', 'door', 'napkin', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the playback device you want to listen to. turn on. press play. enjoyu music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'microwave', 'toothbrush', 'drawing', 'orchid', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'bedroom', 'stovefan', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'mousepad', 'pillow', 'wall', 'pasta', 'oven', 'coffe_maker', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'sauce_pan', 'cupboard']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Cook some pasta on the stove", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pasta>", "[GRAB] <pasta>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <sauce_pan>", "[POUR] <water> <sauce_pan>", "[FIND] <oven>", "[SWITCHON] <oven>", "[PUTBACK] <pasta> <sauce_pan>", "[FIND] <sauce_pan>", "[GRAB] <sauce_pan>", "[PUTBACK] <sauce_pan> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'cpuscreen', 'clothes_underwear', 'hanger', 'phone', 'bedroom', 'bookshelf', 'electric_shaver', 'window', 'towel', 'wall', 'nightstand', 'oven', 'powersocket', 'basket_for_clothes', 'newspaper', 'check', 'oven_mitts', 'after_shave', 'towel_rack', 'mouse', 'kettle', 'dresser', 'laundry_detergent', 'bag', 'table', 'dishwasher', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Pick up my dirty clothes, load the washer, add detergent, turn on washer,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[RUN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'needle', 'floor', 'ceiling', 'computer', 'toaster', 'stovefan', 'mat', 'faucet', 'shower', 'pillow', 'wall', 'oven', 'cup', 'photoframe', 'toilet_paper', 'coffe_maker', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'light', 'creditcard']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "i go to the bathroom. I find the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'band-aids', 'walllamp', 'drawing', 'food_egg', 'bathroom_counter', 'ceiling', 'stamp', 'desk', 'cleaning_solution', 'computer', 'stovefan', 'deck_of_cards', 'faucet', 'tvstand', 'window', 'couch', 'wall', 'nightstand', 'coffe_maker', 'television', 'mirror', 'after_shave', 'paper_towel', 'clothes_dress', 'sink', 'bag', 'bathroom_cabinet', 'dishwasher', 'laser_pointer']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[FIND] <cleaning_solution>', '[RINSE] <cleaning_solution>', '[WIPE] <mirror>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'food_cake', 'stovefan', 'faucet', 'closetdrawer', 'couch', 'form', 'dining_room', 'wall', 'nightstand', 'photoframe', 'television', 'tray', 'doorjamb', 'cards', 'kitchen_counter', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'light', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom.  Find mouthwash or turn on water.  Put mouthwash or water in mouth.  Swish around for 30 seconds.  Spit mouthwash out.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <teeth>", "[RINSE] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'tea_bag', 'bookshelf', 'ceilinglamp', 'dining_room', 'wall', 'oven', 'photoframe', 'bed', 'novel', 'check', 'tray', 'doorjamb', 'towel_rack', 'console', 'pencil', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the Living Room. Look at the Cabinet. Find a Textbook. Grab the textbook. Open the Textbook. Read the textbook.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[FIND] <filing_cabinet>', '[TURNTO] <filing_cabinet>', '[LOOKAT] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'computer', 'stovefan', 'tvstand', 'faucet', 'shower', 'pillow', 'wall', 'cup', 'bed', 'coffe_maker', 'television', 'tray', 'lighter', 'dresser', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Walk to the sink, turn on the faucet, put soap on hands, rub hands together under the water, rinse off hands, turn off faucet, dry hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[WALK] <sink>", "[FIND] <soap>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'ground_coffee', 'detergent', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'bookshelf', 'envelope', 'mat', 'shower', 'faucet', 'homework', 'mousepad', 'wall', 'toy', 'photoframe', 'check', 'coffee_filter', 'food_onion', 'shaving_cream', 'cards', 'mouse', 'dresser', 'hairbrush', 'door', 'soap', 'scrabble']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Enter inside kid's room, begin picking up toys from the floor and putting them in a toy box, leave once once.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'orchid', 'floor', 'bedroom', 'bookshelf', 'toaster', 'tvstand', 'faucet', 'shower', 'stereo', 'closetdrawer', 'couch', 'broom', 'wall', 'wall_clock', 'photoframe', 'bed', 'television', 'mouse', 'kitchen_counter', 'comb', 'dresser', 'sink', 'spoon', 'clothes_skirt']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bath room and go to the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'chair', 'blow_dryer', 'orchid', 'cpuscreen', 'floor', 'phone', 'folder', 'desk', 'bookshelf', 'remote_control', 'computer', 'toaster', 'stovefan', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'photoframe', 'novel', 'television', 'coin', 'coffee_filter', 'doorjamb', 'shoes', 'cupboard', 'sink', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[PUSH] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'toaster', 'mat', 'faucet', 'shower', 'pillow', 'wall', 'nightstand', 'bench', 'toy', 'coffe_maker', 'television', 'curtain', 'after_shave', 'doorjamb', 'towel_rack', 'kitchen_counter', 'chessboard', 'sink', 'home_office', 'bathroom_cabinet', 'table', 'instrument_violin']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[OPEN] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'deck_of_cards', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'door', 'light', 'table']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Keep my keyboard vacuumed and free of debris.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'toaster', 'ceilinglamp', 'window', 'floor_lamp', 'couch', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'comb', 'door', 'bathtub', 'bathroom_cabinet', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'freezer', 'detergent', 'drawing', 'knife', 'wallshelf', 'dvd_player', 'character', 'bookshelf', 'plate', 'shower', 'tvstand', 'dish_soap', 'closetdrawer', 'pillow', 'food_kiwi', 'nightstand', 'powersocket', 'cutting_board', 'photoframe', 'wooden_spoon', 'dishrack', 'dresser', 'cupboard', 'door', 'sink', 'spoon', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <dishrack>', '[PUTBACK] <fork> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishrack>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishrack>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'floor', 'ceiling', 'desk', 'toaster', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'tvstand', 'wall', 'oven', 'box', 'food_food', 'coffe_maker', 'water', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I enter the kitchen and grab a waterglass from the cupboard. I open the fridge and grap some water. I pour the water into the cup and drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[OPEN] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'dining_room', 'wall', 'powersocket', 'oven', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Push power button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'microwave', 'chair', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'computer', 'table_cloth', 'tvstand', 'shower', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'television', 'tray', 'doorjamb', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'board_game']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk into living room, walk to light, turn on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'nail_polish', 'wallshelf', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'piano_bench', 'stovefan', 'mat', 'plate', 'faucet', 'picture', 'tvstand', 'dish_soap', 'food_noodles', 'pillow', 'food_kiwi', 'wall', 'wall_clock', 'oven', 'tape', 'shelf', 'coin', 'tray', 'curtain', 'sink', 'napkin', 'bathroom_cabinet']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Make some dish water, grab a dish rag, put dishes in dish water and wash.,\n    \"actions\": ['[RUN] <sink>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <plate>', '[PUTBACK] <dish_soap> <plate>', '[RINSE] <dish_soap>', '[FIND] <plate>', '[WASH] <plate>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'measuring_cup', 'ground_coffee', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'keyboard', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'oven', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'walllamp', 'bathroom', 'cpuscreen', 'kitchen_cabinet', 'floor', 'tea_bag', 'toaster', 'mat', 'stereo', 'balanceball', 'dish_soap', 'towel', 'wall', 'oven', 'bed', 'light_bulb', 'novel', 'coffe_maker', 'check', 'coin', 'food_donut', 'tray', 'doorjamb', 'towel_rack', 'sink', 'dishwasher', 'vase', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTIN] <dish_soap> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'cat', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'toaster', 'window', 'oil', 'closetdrawer', 'couch', 'form', 'pillow', 'wall', 'wall_clock', 'coffe_maker', 'television', 'curtain', 'mouse', 'home_office', 'table', 'coffee']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'toaster', 'mat', 'ceilinglamp', 'plate', 'stovefan', 'dish_soap', 'broom', 'wall', 'bench', 'food_food', 'coffe_maker', 'tray', 'wooden_spoon', 'towel_rack', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'table', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go the in kitchen then grab the plate on the table. I put the plate and dish soap in the dishwasher. Finally, I turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'scissors', 'tablelamp', 'microwave', 'chair', 'bathroom', 'trashcan', 'drawing', 'clothes_underwear', 'phone', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'closetdrawer', 'candle', 'clothes_pants', 'wall', 'photoframe', 'curtain', 'mouse', 'cupboard', 'clothes_dress', 'soap', 'tooth_paste', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I need to get the dirty clothes basket.  Walk to bathroom.  Put all the clothes in washing machine.  Add soap.  Turn on machine.", "actions": ["[WALK] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Using desktop computer in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'toilet', 'freezer', 'trashcan', 'drawing', 'floor', 'phone', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'toaster', 'mat', 'faucet', 'window', 'wall', 'wall_clock', 'powersocket', 'box', 'cup', 'bed', 'video_game_controller', 'television', 'doorjamb', 'console', 'kitchen_counter', 'door', 'light', 'teeth']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone and pick it up.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'blow_dryer', 'cat', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'ceilinglamp', 'balanceball', 'shower', 'window', 'couch', 'wall', 'powersocket', 'food_food', 'television', 'oven_mitts', 'doorjamb', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'light', 'table', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": My cat looks like it wants attention, so I will go pet it on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'bathroom', 'cat', 'trashcan', 'drawing', 'wallshelf', 'floor', 'desk', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'bench', 'cup', 'food_food', 'television', 'curtain', 'towel_rack', 'mouse', 'sink', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'laptop', 'plate', 'stovefan', 'shower', 'dish_soap', 'wall', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'kettle', 'pencil', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher', 'coffee']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put the dishes on dinner table in the dishwasher and wash them,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[WALK] <table>', '[FIND] <pot>', '[GRAB] <pot>', '[WALK] <dishwasher>', '[PUTBACK] <pot> <dishwasher>', '[CLOSE] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[OPEN] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'desk', 'tea_bag', 'bookshelf', 'ceilinglamp', 'dining_room', 'wall', 'oven', 'photoframe', 'bed', 'novel', 'check', 'tray', 'doorjamb', 'towel_rack', 'console', 'pencil', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to the Living Room. Look at the Cabinet. Find a Textbook. Grab the textbook. Open the Textbook. Read the textbook.", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[FIND] <filing_cabinet>", "[TURNTO] <filing_cabinet>", "[LOOKAT] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink']\nTask:Use computer\nresponse:\n{\n    \"thoughts\": Browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'freezer', 'bathroom', 'nail_polish', 'toothbrush', 'drawing', 'cpuscreen', 'food_egg', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'stovefan', 'ceilinglamp', 'faucet', 'picture', 'powersocket', 'oven', 'bed', 'photoframe', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'food_steak', 'sink', 'door', 'table', 'instrument_violin']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room, walk to the phone reach for the phone and pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'dry_pasta', 'food_ice_cream', 'toilet', 'freezer', 'knifeblock', 'wallshelf', 'folder', 'desk', 'stamp', 'bedroom', 'computer', 'toaster', 'mat', 'floor_lamp', 'dustpan', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'check', 'curtain', 'console', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'nail_polish', 'trashcan', 'drawing', 'orchid', 'floor', 'desk', 'computer', 'food_bacon', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'bench', 'coffe_maker', 'television', 'curtain', 'mouse', 'bathtub', 'home_office', 'table', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_salt', 'toilet', 'chair', 'bathroom', 'trashcan', 'drawing', 'floor', 'character', 'desk', 'purse', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'foundation', 'couch', 'pillow', 'nightstand', 'powersocket', 'novel', 'tray', 'mouse', 'bathroom_cabinet', 'soap', 'teeth']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'food_egg', 'clothes_underwear', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'bench', 'box', 'television', 'curtain', 'towel_rack', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Changing channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[DROP] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'walllamp', 'food_butter', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'mat', 'stereo', 'faucet', 'window', 'mousepad', 'pillow', 'wall', 'powersocket', 'bed', 'coffee_table', 'television', 'doorjamb', 'pencil', 'dresser', 'clothes_dress', 'door', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room and look at the coffee table. Grab the remote control and turn to the television and point the remote control. Put back the remote control.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <coffee_table>', '[TURNTO] <coffee_table>', '[LOOKAT] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['scissors', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'couch', 'pillow', 'wall', 'television', 'wine', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'cupboard', 'controller', 'sink', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'scissors', 'chair', 'freezer', 'knifeblock', 'clothes_underwear', 'bookshelf', 'mop_bucket', 'stovefan', 'shower', 'window', 'couch', 'dining_room', 'nightstand', 'wall_clock', 'powersocket', 'photoframe', 'keys', 'bed', 'towel_rack', 'cd', 'dresser', 'cupboard', 'sink', 'napkin', 'bathroom_cabinet', 'home_office', 'dishwasher', 'board_game']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find the cat.  Pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'microwave', 'toilet', 'freezer', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bookshelf', 'computer', 'toaster', 'mat', 'shower', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'bench', 'cup', 'food_food', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'napkin']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Find groceries. Grab groceries. Walk to refrigerator. Open refrigerator door. Place groceries into refrigerator. Close refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'milk', 'bathroom_counter', 'floor', 'bookshelf', 'table_cloth', 'toaster', 'tvstand', 'shower', 'foundation', 'pillow', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'cup', 'coffe_maker', 'food_onion', 'kitchen_counter', 'dishrack', 'cupboard', 'clothes_dress', 'sink', 'bag', 'table', 'dishwasher', 'dirt', 'shoe_rack']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <dishrack>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <dishrack> <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'phone', 'desk', 'bookshelf', 'toaster', 'ceilinglamp', 'shower', 'faucet', 'couch', 'dining_room', 'nightstand', 'wall_clock', 'oven', 'wall', 'bed', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the water, wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <water>", "[FIND] <water>", "[TURNTO] <water>", "[LOOKAT] <water>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'remote_control', 'computer', 'button', 'toaster', 'mat', 'shower', 'food_peanut_butter', 'food_kiwi', 'nightstand', 'oven', 'photoframe', 'television', 'oven_mitts', 'tray', 'kettle', 'cards', 'pencil', 'kitchen_counter', 'towel_rack', 'bathtub', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk into living room. Find television remote control. Pick up remote control. Press button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mousepad', 'bowl', 'wall', 'newspaper', 'bed', 'novel', 'curtain', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "go to the room, sit in the chair, turn on the computer, look for the news", "actions": ["[WALK] <bedroom>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'microwave', 'chair', 'freezer', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'stereo', 'picture', 'mat', 'closetdrawer', 'couch', 'dining_room', 'wall_clock', 'bench', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'home_office', 'instrument_guitar']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the living room and then I walk to the light switch in there and turn the lights on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'tablelamp', 'microwave', 'toilet', 'trashcan', 'hanger', 'bathroom_counter', 'bedroom', 'keyboard', 'stovefan', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'nightstand', 'wall', 'oven', 'wall_clock', 'coffe_maker', 'television', 'oven_mitts', 'curtain', 'lighter', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'light', 'instrument_guitar', 'teeth']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <home_office>", "[DROP] <plate>", "[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <home_office>", "[DROP] <cup>", "[WALK] <dining_room>", "[WALK] <fork>", "[FIND] <fork>", "[GRAB] <fork>", "[WALK] <home_office>", "[DROP] <fork>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'toaster', 'window', 'mousepad', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'wall_clock', 'television', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'clothes_shirt']\nTask:Organize closet\nresponse:\n{\n    \"thoughts\": Organize the hanging clothes in my closet by color and type and get rid of clothes I don't wear.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PULL] <clothes_pants>', '[PUTIN] <clothes_pants> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUSH] <clothes_shirt>', '[PUTIN] <clothes_shirt> <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[CLOSE] <dresser>', '[WALK] <basket_for_clothes>', '[PUTBACK] <clothes_shirt> <basket_for_clothes>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'chair', 'bathroom', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'phone', 'character', 'computer', 'toaster', 'mat', 'tvstand', 'couch', 'clothes_hat', 'pillow', 'wall', 'wall_clock', 'bench', 'powersocket', 'cup', 'bed', 'food_food', 'television', 'mouse', 'pencil', 'bathroom_cabinet', 'table']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "walk to entrance hall. find a table. pull table. walk to living rook. push table.", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[PULL] <table>", "[WALK] <home_office>", "[WALK] <table>", "[PUSH] <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'food_kiwi', 'wall', 'nightstand', 'oven', 'photoframe', 'television', 'doorjamb', 'headset', 'dresser', 'bathroom_cabinet', 'soap']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I find the remote then I push the button to switch it to the tv channel i want to watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'chair', 'walllamp', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'character', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'toaster', 'plate', 'dish_soap', 'mousepad', 'closetdrawer', 'oven', 'powersocket', 'bed', 'photoframe', 'check', 'water', 'cards', 'hairbrush', 'sink', 'bathroom_cabinet', 'light']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen, fins soap, arrange dishes, wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[FIND] <dish_soap>", "[FIND] <water>", "[GRAB] <plate>", "[GRAB] <dish_soap>", "[WASH] <plate>", "[RINSE] <plate>", "[DROP] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bedroom', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'food_food', 'television', 'wooden_spoon', 'curtain', 'diary', 'dresser', 'cupboard', 'clothes_dress', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": You walk to the living room and sit on the sofa. You grab the remote control, look at it and point at the television, you watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'tablelamp', 'freezer', 'trashcan', 'dough', 'cpuscreen', 'ceiling', 'bedroom', 'keyboard', 'toaster', 'mat', 'plate', 'shower', 'faucet', 'dish_soap', 'mousepad', 'closetdrawer', 'dining_room', 'wall_clock', 'bench', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'clothes_dress', 'sink', 'home_office', 'dishwasher', 'coffee']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the  kitchen. Open the dishwasher. Put plates into the dishwasher. Put dish soap into the dishwasher. Close the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'crayon', 'toilet', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'mail', 'food_bacon', 'clothes_jacket', 'table_cloth', 'faucet', 'mousepad', 'closetdrawer', 'food_jam', 'nightstand', 'powersocket', 'novel', 'food_food', 'centerpiece', 'diary', 'console', 'cupboard', 'bathroom_cabinet', 'table', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Pick up the ringing phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[POINTAT] <table>', '[FIND] <phone>', '[GRAB] <phone>', '[PUTOBJBACK] <phone>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'floor_lamp', 'mousepad', 'couch', 'dining_room', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'doorjamb', 'cd', 'door', 'bathtub', 'home_office', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to wall, find light switch, turn on lights", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'freezer', 'orchid', 'drawing', 'wallshelf', 'floor', 'phone', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'shower', 'window', 'mousepad', 'pillow', 'nightstand', 'wall', 'oven', 'wall_clock', 'bed', 'novel', 'curtain', 'towel_rack', 'cupboard', 'door', 'sink', 'light', 'laser_pointer']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the bedroom. Grab the book out of the nightstand. Lay down in bed. Open the book. Read the book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <nightstand>', '[FIND] <nightstand>', '[OPEN] <nightstand>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[TURNTO] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'bookshelf', 'clothes_jacket', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'shower', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and find the sofa. I sit on the sofa and get comfortable.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'stamp', 'bedroom', 'stovefan', 'toaster', 'plate', 'faucet', 'shower', 'dish_soap', 'food_jam', 'towel', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'ice', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'orchid', 'floor', 'desk', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'pillow', 'nightstand', 'wall_clock', 'oven', 'wall', 'coffe_maker', 'mouthwash', 'television', 'tray', 'towel_rack', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to phone, grab phone, pick up phone, put to ear", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_cake', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'couch', 'wall', 'powersocket', 'bed', 'television', 'doorjamb', 'mouse', 'hairdryer', 'door', 'sink', 'home_office', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Using desktop computer in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['tea', 'toilet', 'dog', 'walllamp', 'trashcan', 'orchid', 'drawing', 'bedroom', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'bench', 'basket_for_clothes', 'tray', 'wooden_spoon', 'kitchen_counter', 'clothes_dress', 'hairdryer', 'sink', 'light', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the washing machine. Open the washing machine. Find the basket for clothes. Find the clothes_dress. Grab the clothes_dress. Put the clothes_dress into the washing machine. Find the clothes_shirt. Grab the clothes_shirt. Put the clothes_shirt into the washing machine. Close the washing machine. Switch on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <basket_for_clothes>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'crayon', 'tea', 'freezer', 'walllamp', 'bathroom', 'detergent', 'blow_dryer', 'drawing', 'floor', 'desk', 'thread', 'mop_bucket', 'tvstand', 'towel', 'wall', 'toy', 'cup', 'check', 'razor', 'food_donut', 'wine', 'centerpiece', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'laundry_detergent', 'light', 'soap']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Robot has a vacuum underneath its feat/body that will allow it to clean the surface of the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[WALK] <dining_room>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[WALK] <home_office>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <filing_cabinet>', '[PUTBACK] <vacuum_cleaner> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'window', 'mat', 'closetdrawer', 'wall', 'wall_clock', 'bed', 'photoframe', 'doorjamb', 'console', 'kitchen_counter', 'door', 'hairdryer', 'sink', 'light', 'clothes_skirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on light in an area", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_ice_cream', 'toothbrush', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'alcohol', 'brush', 'shoe-shine_kit', 'faucet', 'mousepad', 'closetdrawer', 'candle', 'clothes_pants', 'wall', 'standingmirror', 'food_food', 'television', 'tray', 'coffee_pot', 'towel_rack', 'dresser', 'bathtub', 'bathroom_cabinet', 'vase']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Using soft brush, carefully dust both Indonesian wedding doll statues.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <brush>', '[GRAB] <brush>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['chair', 'drawing', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'dvd_player', 'bookshelf', 'remote_control', 'button', 'piano_bench', 'ceilinglamp', 'tvstand', 'mat', 'wall', 'wall_clock', 'bench', 'oven', 'nightstand', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'cd', 'door', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control", "actions": ["[WALK] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'freezer', 'blow_dryer', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'shoe-shine_kit', 'couch', 'microphone', 'food_jam', 'wall', 'coffe_maker', 'check', 'television', 'curtain', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'colander', 'tooth_paste']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the television. Walk to the television. Switch on the television. Find the sofa. Walk to the sofa. Sit on the sofa. Find the remote control. Grab the remote control. Look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[WALK] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'chair', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'table_cloth', 'toaster', 'stovefan', 'tvstand', 'shower', 'faucet', 'closetdrawer', 'nightstand', 'wall', 'bench', 'coffe_maker', 'check', 'television', 'tray', 'pencil', 'shoes', 'sink', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "turn radio on, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'shoe_rack', 'food_salt', 'mop', 'microwave', 'dog', 'jelly', 'walllamp', 'drawing', 'needle', 'food_egg', 'desk', 'stamp', 'computer', 'ceilinglamp', 'window', 'mousepad', 'food_orange', 'pillow', 'bench', 'cup', 'oven_mitts', 'wooden_spoon', 'pencil', 'dresser', 'maindoor', 'bag', 'table', 'instrument_violin', 'creditcard']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": go to radio. run on radio.,\n    \"actions\": ['[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'scissors', 'tea', 'freezer', 'walllamp', 'wallshelf', 'hanger', 'kitchen_cabinet', 'floor', 'bathroom_counter', 'bedroom', 'bookshelf', 'ceilinglamp', 'dish_soap', 'dining_room', 'nightstand', 'wall_clock', 'bed', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'door', 'clothes_scarf', 'bathtub', 'light', 'dishwasher', 'table', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTIN] <dish_soap> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'keyboard', 'mail', 'mat', 'faucet', 'wall', 'nightstand', 'powersocket', 'oven', 'cup', 'photoframe', 'bed', 'novel', 'curtain', 'doorjamb', 'diary', 'cupboard', 'door', 'sink', 'instrument_guitar']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk over to the bookshelf. Pick storybook off the shelf. Sit on bed. Open storybook and read out loud.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[WALK] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'powersocket', 'oven', 'box', 'television', 'ice', 'curtain', 'pencil', 'kitchen_counter', 'door', 'hairdryer', 'instrument_guitar', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "allow sofa to support weight in order to decrease gravity induced entropy", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'blender', 'beer', 'toilet', 'jelly', 'freezer', 'blow_dryer', 'detergent', 'orchid', 'knifeblock', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'mail', 'computer', 'bookmark', 'stereo', 'window', 'couch', 'microphone', 'check', 'television', 'rag', 'pencil', 'controller', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Pet Cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'food_salt', 'toilet', 'chair', 'bathroom', 'trashcan', 'drawing', 'floor', 'character', 'desk', 'purse', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'foundation', 'couch', 'pillow', 'nightstand', 'powersocket', 'novel', 'tray', 'mouse', 'bathroom_cabinet', 'soap', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'tablelamp', 'microwave', 'detergent', 'clothes_underwear', 'ceiling', 'character', 'desk', 'mail', 'table_cloth', 'bookmark', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'mousepad', 'candle', 'food_jam', 'food_orange', 'nightstand', 'powersocket', 'newspaper', 'coin', 'food_onion', 'pencil', 'dresser', 'home_office', 'dishwasher', 'tooth_paste', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'chair', 'walllamp', 'toothbrush', 'wallshelf', 'dvd_player', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'mousepad', 'couch', 'form', 'pillow', 'bowl', 'nightstand', 'wall_clock', 'bench', 'oven', 'powersocket', 'clothes_gloves', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'hairdryer', 'bathtub', 'home_office', 'table']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to my bed room and sit on the chair and in front of me is my desktop computer and start writing an email.", "actions": ["[WALK] <bathroom>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'band-aids', 'toothbrush', 'orchid', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'faucet', 'picture', 'shower', 'mousepad', 'candle', 'food_orange', 'oven', 'basket_for_clothes', 'bed', 'check', 'food_food', 'television', 'tray', 'after_shave', 'headset', 'comb', 'dresser', 'laundry_detergent', 'napkin', 'bathroom_cabinet']\nTask:Sit in chair\nresponse:\n{\n    \"thoughts\": I sit in a chair to rest and eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'tea', 'trashcan', 'drawing', 'orchid', 'floor', 'character', 'desk', 'bookshelf', 'computer', 'toaster', 'mat', 'stovefan', 'tvstand', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'wall_clock', 'cup', 'bed', 'food_donut', 'curtain', 'towel_rack', 'sink', 'napkin', 'bathtub']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <sheets>", "[PULL] <sheets>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'homework', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'keys', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to Office,Walk to desk ,sit down in chair, turn on computer, start homework.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <check>', '[FIND] <homework>', '[TURNTO] <homework>', '[LOOKAT] <homework>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'band-aids', 'walllamp', 'drawing', 'wallshelf', 'floor', 'character', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'bench', 'photoframe', 'television', 'tray', 'curtain', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'home_office', 'table', 'clothes_shirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Sitting down and relaxing on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'freezer', 'blow_dryer', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'shoe-shine_kit', 'couch', 'microphone', 'food_jam', 'wall', 'coffe_maker', 'check', 'television', 'curtain', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'colander', 'tooth_paste']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the television. Walk to the television. Switch on the television. Find the sofa. Walk to the sofa. Sit on the sofa. Find the remote control. Grab the remote control. Look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[WALK] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'faucet', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'television', 'doorjamb', 'mouse', 'door', 'sink', 'light', 'instrument_violin']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I enter the bedroom and find the light. Then I turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'microwave', 'toilet', 'freezer', 'ground_coffee', 'longboard', 'shampoo', 'character', 'thread', 'food_bacon', 'piano_bench', 'balanceball', 'stereo', 'window', 'picture', 'mousepad', 'pillow', 'toy', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'oven_mitts', 'colander', 'doorjamb', 'shaving_cream', 'dresser', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[WALK] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'tvstand', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'food_jam', 'wall', 'nightstand', 'oven', 'newspaper', 'food_food', 'vacuum_cleaner', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'band-aids', 'freezer', 'nail_polish', 'bathroom', 'cat', 'orchid', 'cpuscreen', 'phone', 'character', 'keyboard', 'mat', 'faucet', 'wall', 'oven', 'bed', 'keys', 'coffe_maker', 'mouthwash', 'food_onion', 'curtain', 'centerpiece', 'console', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'milk', 'wallshelf', 'cpuscreen', 'hanger', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'table_cloth', 'stovefan', 'tvstand', 'window', 'closetdrawer', 'candle', 'dining_room', 'wall_clock', 'oven', 'bench', 'photoframe', 'television', 'tray', 'kitchen_counter', 'door', 'bathtub', 'sink', 'table', 'soap', 'creditcard']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'plate', 'faucet', 'tvstand', 'clothes_hat', 'pillow', 'wall', 'oven', 'powersocket', 'newspaper', 'photoframe', 'check', 'coffe_maker', 'doorjamb', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'clothes_scarf', 'table', 'clothes_skirt', 'dishwasher']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[TURNTO] <plate>', '[LOOKAT] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WASH] <plate>', '[PUTBACK] <plate> <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'phone', 'bedroom', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'curtain', 'rag', 'cupboard', 'table', 'colander']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I go to the living room and sit in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'chair', 'freezer', 'detergent', 'bathroom_counter', 'desk', 'bedroom', 'keyboard', 'button', 'stovefan', 'ceilinglamp', 'stereo', 'shower', 'window', 'clothes_socks', 'dining_room', 'nightstand', 'wall', 'bench', 'oven', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'dresser', 'napkin', 'bathroom_cabinet', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room, look for radio, grab on button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'dog', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'form', 'wall', 'powersocket', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'mouse', 'cupboard', 'bathtub', 'sink', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "go to home office. turn the computer on. begin working", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'chair', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'stereo', 'floor_lamp', 'mousepad', 'wall', 'nightstand', 'novel', 'video_game_controller', 'television', 'mouse', 'door', 'napkin']\nTask:Study\nresponse:\n{\n    \"thoughts\": Study norwegian,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'floor', 'character', 'keyboard', 'bookshelf', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'wall_clock', 'toilet_paper', 'mouthwash', 'television', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, find sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'freezer', 'walllamp', 'food_carrot', 'desk', 'food_cake', 'remote_control', 'button', 'ceilinglamp', 'faucet', 'stereo', 'window', 'mousepad', 'couch', 'wall_clock', 'bench', 'box', 'cup', 'bed', 'keys', 'newspaper', 'food_food', 'ice', 'after_shave', 'maindoor', 'bathroom_cabinet', 'light', 'vase', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[PUSH] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'wall', 'powersocket', 'food_food', 'television', 'curtain', 'doorjamb', 'mouse', 'cupboard', 'door', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Turn on computer, pull up browser, browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'glue', 'jelly', 'freezer', 'detergent', 'food_carrot', 'orchid', 'clothes_underwear', 'hanger', 'floor', 'bedroom', 'thread', 'bookmark', 'mat', 'stereo', 'shower', 'mousepad', 'oil', 'nightstand', 'light_bulb', 'coffe_maker', 'television', 'tray', 'pencil', 'shoes', 'bathroom_cabinet', 'dishwasher', 'colander', 'dirt']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Go to the living room.  Find the telephone.  Lift up the receiver.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'brush', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'wall', 'powersocket', 'bench', 'coffe_maker', 'mouthwash', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'kettle', 'cupboard', 'sink', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "first i will walk into my home office and i needed to know the today's gold prices so wih the help of keyboard and mouse i will browse and find it", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'towel', 'wall', 'wall_clock', 'oven', 'television', 'tray', 'rag', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <bathroom_cabinet>', '[FIND] <shower>', '[RINSE] <shower>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'walllamp', 'nail_polish', 'orchid', 'cpuscreen', 'phone', 'folder', 'bedroom', 'desk', 'keyboard', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'floor_lamp', 'iron', 'mousepad', 'couch', 'powersocket', 'photoframe', 'coffe_maker', 'curtain', 'towel_rack', 'pencil', 'cupboard', 'bathtub', 'napkin', 'light', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the living room, walk to the lamp, switch on the lamp.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'juice', 'food_ice_cream', 'bathroom', 'orchid', 'floor', 'ceiling', 'phone', 'shampoo', 'bookshelf', 'tvstand', 'stereo', 'shower', 'couch', 'wall', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'food_food', 'television', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'controller', 'home_office', 'light', 'coffee']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": first i will go to home office and switch on the lights and answer the call after then i will put back it and switch off the lights and close the door,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <phone>', '[GRAB] <phone>', '[PUTOBJBACK] <phone>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'dvd_player', 'desk', 'bedroom', 'bookshelf', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'tray', 'curtain', 'controller', 'sink', 'bathroom_cabinet', 'table', 'home_office', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music", "actions": ["[WALK] <home_office>", "[WALK] <dvd_player>", "[FIND] <dvd_player>", "[TURNTO] <dvd_player>", "[LOOKAT] <dvd_player>", "[SWITCHON] <dvd_player>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['cat', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'iron', 'towel', 'wall', 'wall_clock', 'oven', 'nightstand', 'tape', 'novel', 'food_food', 'television', 'tray', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'food_salt', 'ground_coffee', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'food_cake', 'ceilinglamp', 'mat', 'stereo', 'window', 'clothes_socks', 'wall', 'bench', 'cup', 'bed', 'coffe_maker', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk in to the living room. I find the light switch. I turn on the lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'band-aids', 'drawing', 'longboard', 'floor', 'ceiling', 'folder', 'computer', 'clothes_jacket', 'alcohol', 'ceilinglamp', 'balanceball', 'window', 'foundation', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'television', 'wine', 'towel_rack', 'pencil', 'shoes', 'dresser', 'cupboard', 'table', 'light', 'dishwasher']\nTask:Arrange furniture\nresponse:\n{\n    \"thoughts\": Move furniture where you want it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[PUSH] <coffee_table>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'food_salt', 'toilet', 'bills', 'hanger', 'bathroom_counter', 'floor', 'character', 'stovefan', 'tvstand', 'faucet', 'mousepad', 'towel', 'dining_room', 'wall', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'water', 'tray', 'curtain', 'doorjamb', 'sink', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Walk to sink and turn on the water. Grab the soap and wash the hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <water>", "[TURNTO] <water>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'toilet', 'jelly', 'walllamp', 'trashcan', 'knifeblock', 'knife', 'desk', 'keyboard', 'computer', 'button', 'brush', 'window', 'mousepad', 'couch', 'bench', 'cutting_board', 'novel', 'check', 'coffe_maker', 'television', 'toothbrush_holder', 'food_rice', 'curtain', 'headset', 'dresser', 'home_office', 'table', 'dishwasher']\nTask:Lock door\nresponse:\n{\n    \"thoughts\": Turn the lock.,\n    \"actions\": ['[WALK] <bedroom>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'bathroom', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'folder', 'bedroom', 'computer', 'stovefan', 'window', 'closetdrawer', 'powersocket', 'basket_for_clothes', 'clothes_gloves', 'food_food', 'doorjamb', 'kettle', 'mouse', 'shoes', 'dresser', 'cupboard', 'sink', 'home_office', 'table', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'envelope', 'computer', 'mop_bucket', 'stovefan', 'mat', 'tvstand', 'toaster', 'foundation', 'mousepad', 'pillow', 'wall', 'standingmirror', 'photoframe', 'toilet_paper', 'oven_mitts', 'tray', 'mouse', 'sink']\nTask:Read book\nresponse:\n{\n    \"thoughts\": go to living room,find the text book on the self ,sat on the chair ,open and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'mop_bucket', 'stovefan', 'ceilinglamp', 'tvstand', 'homework', 'mousepad', 'couch', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'television', 'doorjamb', 'mouse', 'headset', 'cupboard', 'table', 'shoe_rack']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to the home office to find the computer. I turn switch computer on , i open the email. i write the email and  send it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'computer', 'button', 'brush', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'standingmirror', 'coffe_maker', 'television', 'food_rice', 'wine', 'curtain', 'doorjamb', 'diary', 'pencil', 'door', 'bathroom_cabinet', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['chair', 'bathroom', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'stamp', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'dining_room', 'wall', 'powersocket', 'oven', 'bed', 'newspaper', 'television', 'doorjamb', 'kitchen_counter', 'chessboard', 'dresser', 'controller', 'door', 'bathroom_cabinet', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into the bedroom and walk to the light switch. Flip the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'tape', 'photoframe', 'television', 'doorjamb', 'kettle', 'mouse', 'dresser']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find the television, switch on the television, find the sofa, walk to sofa, sit in sofa, watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'bathroom', 'drawing', 'floor', 'character', 'bedroom', 'stamp', 'computer', 'mat', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'closetdrawer', 'clothes_hat', 'pillow', 'dining_room', 'broom', 'wall', 'bed', 'doorjamb', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'home_office', 'board_game', 'dirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room ,find electrical outlet , switch on eletrical outlet", "actions": ["[WALK] <home_office>", "[WALK] <electrical_outlet>", "[FIND] <electrical_outlet>", "[SWITCHON] <electrical_outlet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'bathroom_counter', 'phone', 'tea_bag', 'keyboard', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'brush', 'foundation', 'window', 'closetdrawer', 'couch', 'broom', 'wall_clock', 'toy', 'food_cheese', 'novel', 'food_donut', 'tray', 'kitchen_counter', 'dresser', 'bathtub', 'light', 'shoe_rack']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to the living room, Find Phone stand. Find Phone. Pickup the receiver to attend the call.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'beer', 'trashcan', 'dough', 'cpuscreen', 'desk', 'keyboard', 'toaster', 'stovefan', 'tvstand', 'shower', 'faucet', 'window', 'oil', 'couch', 'dining_room', 'cutting_board', 'bed', 'food_food', 'television', 'ice', 'doorjamb', 'towel_rack', 'cd', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I hear the phone ringing in the living room. I walk to the living room and go to the phone. I pick it up and answer the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>", "[SWITCHON] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'oven', 'toy', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'pencil', 'dresser', 'sink', 'board_game']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to living room, walk to TV, grab TV remote, turn on TV, put down the remote, walk to sofa, sit in soft, watch TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'band-aids', 'trashcan', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'envelope', 'clothes_jacket', 'mat', 'shower', 'stereo', 'microphone', 'wall', 'nightstand', 'bed', 'photoframe', 'food_food', 'rag', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'nail_polish', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'bedroom', 'tea_bag', 'laptop', 'tvstand', 'clothes_socks', 'closetdrawer', 'couch', 'clothes_hat', 'wall', 'nightstand', 'powersocket', 'photoframe', 'coffe_maker', 'check', 'television', 'doorjamb', 'towel_rack', 'cards', 'dresser', 'door', 'home_office', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk into the dining room and turn on the lights,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'freezer', 'ground_coffee', 'detergent', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'stovefan', 'toaster', 'faucet', 'shower', 'closetdrawer', 'candle', 'dining_room', 'wall', 'oven', 'photoframe', 'bed', 'coffe_maker', 'coffee_filter', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'table']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Make coffee in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'toothbrush', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'shampoo', 'floor', 'ceiling', 'desk', 'mop_bucket', 'stovefan', 'laptop', 'tvstand', 'faucet', 'wall', 'nightstand', 'oven', 'tape', 'food_food', 'tray', 'curtain', 'kettle', 'kitchen_counter', 'cupboard', 'conditioner', 'sink', 'bathroom_cabinet', 'table']\nTask:Wipe down counter\nresponse:\n{\n    \"thoughts\": Clean counter,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <rag>', '[SQUEEZE] <rag>', '[SWITCHOFF] <faucet>', '[WALK] <kitchen_counter>', '[WIPE] <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'trashcan', 'toothbrush', 'orchid', 'wallshelf', 'hanger', 'floor', 'phone', 'keyboard', 'toaster', 'tvstand', 'shower', 'plate', 'oil', 'closetdrawer', 'pillow', 'tape', 'basket_for_clothes', 'photoframe', 'check', 'television', 'toothbrush_holder', 'shaving_cream', 'mouse', 'towel_rack', 'kitchen_counter', 'sink', 'bathtub', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, take toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now brush teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[TURNTO] <toothbrush>", "[LOOKAT] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'sheets', 'scissors', 'microwave', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'toothbrush', 'floor', 'phone', 'toaster', 'mat', 'tvstand', 'dustpan', 'wall', 'box', 'photoframe', 'coffe_maker', 'check', 'oven_mitts', 'food_rice', 'curtain', 'coffee_pot', 'kitchen_counter', 'cupboard', 'hairdryer', 'dishwasher', 'soap', 'coffee']\nTask:Take dishes out of dishwasher\nresponse:\n{\n    \"thoughts\": Take clean dishes out of the dishwasher and put away,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[CLOSE] <dishwasher>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'mat', 'stereo', 'faucet', 'mousepad', 'wall', 'wall_clock', 'cup', 'bed', 'photoframe', 'novel', 'coffe_maker', 'television', 'mouse', 'kitchen_counter', 'headset', 'cupboard', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to my bedroom.  I locate my stereo and turn it on.  I put on my headphones and sit to listen.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[WALK] <stereo>", "[SWITCHON] <stereo>", "[FIND] <headset>", "[GRAB] <headset>", "[PUTON] <headset>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'walllamp', 'milk', 'clothes_underwear', 'character', 'bedroom', 'keyboard', 'clothes_socks', 'candle', 'clothes_pants', 'wall', 'nightstand', 'powersocket', 'basket_for_clothes', 'photoframe', 'bed', 'food_food', 'curtain', 'diary', 'shoes', 'headset', 'laundry_detergent', 'sink', 'home_office', 'table', 'colander', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[DROP] <basket_for_clothes>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[GRAB] <laundry_detergent>', '[PUTOBJBACK] <laundry_detergent>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'table_cloth', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'tvstand', 'wall', 'wall_clock', 'oven', 'powersocket', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen counter, grab a grocery, walk to fridge, open fridge, put the grocery into fridge, and close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'floor', 'alcohol', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'deck_of_cards', 'window', 'shower', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'photoframe', 'keys', 'television', 'curtain', 'doorjamb', 'mouse', 'home_office', 'table', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into bedrrom and to the sofa. Sit on the sofa and touch cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'bathroom', 'orchid', 'wallshelf', 'floor', 'ceiling', 'character', 'keyboard', 'piano_bench', 'ceilinglamp', 'plate', 'faucet', 'window', 'dish_soap', 'wall_clock', 'bed', 'coffe_maker', 'coffee_filter', 'tray', 'towel_rack', 'kitchen_counter', 'dresser', 'door', 'napkin', 'home_office', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Put dishes in the dishwasher and start the dishwasher.", "actions": ["[RUN] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[WALK] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'plate', 'iron', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'cutting_board', 'clothes_gloves', 'check', 'coffe_maker', 'food_food', 'curtain', 'food_onion', 'mouse', 'comb', 'dresser', 'clothes_dress', 'sink', 'light', 'coffee']\nTask:Hang up jacket\nresponse:\n{\n    \"thoughts\": I put clothes on hangers and hang them in the closet,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <clothes_dress>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[FIND] <hanger>', '[OPEN] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[WALK] <dresser>', '[PUTIN] <clothes_dress> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'bathroom', 'trashcan', 'drawing', 'orchid', 'floor', 'phone', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'faucet', 'picture', 'window', 'closetdrawer', 'couch', 'dining_room', 'bed', 'light_bulb', 'keys', 'check', 'video_game_controller', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'home_office']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'milk', 'drawing', 'cpuscreen', 'shampoo', 'folder', 'bookshelf', 'computer', 'table_cloth', 'ceilinglamp', 'mat', 'oil', 'mousepad', 'couch', 'towel', 'nightstand', 'wall', 'cutting_board', 'cup', 'photoframe', 'bed', 'food_donut', 'mouse', 'kitchen_counter', 'pajamas', 'sink', 'napkin', 'home_office', 'tooth_paste']\nTask:Open front door\nresponse:\n{\n    \"thoughts\": Walk over to the front door, turn the knob, and open the door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'chair', 'freezer', 'walllamp', 'food_butter', 'bills', 'cpuscreen', 'bathroom_counter', 'phone', 'ceiling', 'character', 'bedroom', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'floor_lamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'oven', 'cup', 'photoframe', 'television', 'kitchen_counter', 'dresser', 'bathroom_cabinet', 'table', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch, flip the switch to the on position", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'food_pizza', 'toilet', 'chair', 'bathroom_counter', 'floor', 'ceiling', 'cleaning_solution', 'food_bacon', 'mop_bucket', 'toaster', 'stovefan', 'faucet', 'shower', 'mousepad', 'dustpan', 'pillow', 'wall', 'nightstand', 'powersocket', 'coffe_maker', 'console', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'pajamas', 'sink', 'napkin', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Washing hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[SCRUB] <hands_both>', '[SWITCHON] <faucet>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['scissors', 'microwave', 'toilet', 'walllamp', 'drawing', 'bills', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'television', 'tray', 'curtain', 'towel_rack', 'pencil', 'kitchen_counter', 'bathtub', 'spoon', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Draft home\nresponse:\n{\n    \"thoughts\": Using mathematics and CAD software design a home,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'juice', 'microwave', 'toilet', 'walllamp', 'orchid', 'bathroom_counter', 'bookshelf', 'bookmark', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'wall_clock', 'food_food', 'coffe_maker', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'kitchen_counter', 'cupboard', 'bathtub', 'home_office', 'bathroom_cabinet']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'filing_cabinet', 'food_bread', 'tablelamp', 'mop', 'toilet', 'freezer', 'walllamp', 'desk', 'keyboard', 'toaster', 'ceilinglamp', 'laptop', 'mousepad', 'dining_room', 'nightstand', 'cup', 'photoframe', 'clothes_gloves', 'novel', 'coffe_maker', 'curtain', 'dresser', 'controller', 'clothes_dress', 'door', 'sink', 'light', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Find a cat in the kitchen and touch it,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'walllamp', 'cat', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'envelope', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'television', 'curtain', 'hair', 'bathroom_cabinet', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I would walk into the living room. Then sit on the couch and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'ground_coffee', 'food_butter', 'dough', 'drawing', 'floor', 'desk', 'bookmark', 'mat', 'faucet', 'floor_lamp', 'window', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'novel', 'television', 'curtain', 'doorjamb', 'chessboard', 'controller', 'maindoor', 'home_office', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": First I walk into the living room, I find the lamp and switch it on. Then I find a book, grab it, sit in the sofa, open the book and read it. When I'm finished, I find a bookmark, grab it and put it in the book. Then I close the book, put it back, and switch off the lamp.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>', '[SWITCHOFF] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'bedroom', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'shower', 'mousepad', 'couch', 'pillow', 'wall', 'check', 'television', 'curtain', 'doorjamb', 'dresser', 'door', 'table', 'colander']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'blow_dryer', 'milk', 'bills', 'orchid', 'floor', 'ceiling', 'character', 'food_cake', 'alcohol', 'stovefan', 'homework', 'closetdrawer', 'pillow', 'broom', 'wall', 'oven', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'water', 'tray', 'kitchen_counter', 'shoes', 'sink', 'light', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['cd_player', 'walllamp', 'needle', 'cpuscreen', 'music_stand', 'floor', 'phone', 'ceiling', 'bedroom', 'tea_bag', 'desk', 'toaster', 'mat', 'window', 'mousepad', 'closetdrawer', 'couch', 'food_peanut_butter', 'pillow', 'nightstand', 'bench', 'photoframe', 'check', 'food_food', 'curtain', 'mouse', 'kitchen_counter', 'door', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to the living and turn to the music stand. Switch on the CD player to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <music_stand>", "[FIND] <music_stand>", "[TURNTO] <music_stand>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_carrot', 'drawing', 'floor', 'character', 'computer', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'candle', 'wall', 'wall_clock', 'oven', 'cup', 'photoframe', 'bed', 'novel', 'check', 'toilet_paper', 'television', 'curtain', 'towel_rack', 'mouse', 'hairbrush', 'laundry_detergent', 'table', 'clothes_skirt']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the couch in the living room, sit on ti and read the book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'toilet', 'chair', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'bedroom', 'bookshelf', 'computer', 'mat', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'nightstand', 'oven', 'coffe_maker', 'television', 'cards', 'towel_rack', 'cupboard', 'napkin', 'bathroom_cabinet']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the sink. Take dishes out of sink. Open dishwasher and place dishes in dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to over to desk, sit down on chair and face the computer, turn the computer on, use arms to place hands on the keyboard, type work into the computer and use eyes to review.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'drawing', 'orchid', 'floor', 'bedroom', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'floor_lamp', 'window', 'closetdrawer', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'wall_clock', 'clothes_gloves', 'novel', 'television', 'curtain', 'towel_rack', 'mouse', 'comb', 'cupboard', 'sink', 'table', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "First I walk into the living room, I find the lamp and switch it on. Then I find a book, grab it, sit in the sofa, open the book and read it. When I'm finished, I find a bookmark, grab it and put it in the book. Then I close the book, put it back, and switch off the lamp.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[PUTBACK] <bookmark> <novel>", "[PUTOBJBACK] <novel>", "[SWITCHOFF] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'dry_pasta', 'food_ice_cream', 'toilet', 'freezer', 'knifeblock', 'wallshelf', 'folder', 'desk', 'stamp', 'bedroom', 'computer', 'toaster', 'mat', 'floor_lamp', 'dustpan', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'check', 'curtain', 'console', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'freezer', 'walllamp', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'microphone', 'pillow', 'wall', 'wall_clock', 'nightstand', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'home_office', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room and sit on the couch, ease into the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'stamp', 'bedroom', 'stovefan', 'toaster', 'plate', 'faucet', 'shower', 'dish_soap', 'food_jam', 'towel', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'ice', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'thread', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'powersocket', 'bench', 'keys', 'television', 'doorjamb', 'diary', 'mouse', 'comb', 'cupboard', 'bathtub', 'home_office', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Browsing the world wide web for entertainment", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bills', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'television', 'toothbrush_holder', 'wine', 'doorjamb', 'mouse', 'dresser', 'hairbrush', 'controller', 'clothes_dress', 'home_office', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'bedroom', 'piano_bench', 'stovefan', 'toaster', 'plate', 'tvstand', 'faucet', 'shower', 'ceilinglamp', 'dish_soap', 'pillow', 'wall', 'bed', 'coffe_maker', 'television', 'tray', 'kettle', 'kitchen_counter', 'dishrack', 'sink', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>", "[GRAB] <plate>", "[WIPE] <plate>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[PUTOBJBACK] <dishrack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'walllamp', 'bathroom', 'ceilingfan', 'trashcan', 'drawing', 'floor', 'ceiling', 'food_bacon', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'couch', 'wall', 'nightstand', 'wall_clock', 'bench', 'bed', 'food_food', 'television', 'coin', 'curtain', 'mouse', 'controller', 'table', 'vase', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sitting on the couch watching my favorite television show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'ground_coffee', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'stovefan', 'toaster', 'faucet', 'shower', 'window', 'closetdrawer', 'pillow', 'wall', 'oven', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'table']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'walllamp', 'milk', 'floor', 'keyboard', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'faucet', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'toy', 'basket_for_clothes', 'coffe_maker', 'video_game_controller', 'tray', 'doorjamb', 'kettle', 'dresser', 'cupboard', 'clothes_dress', 'sink', 'table']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'couch', 'candle', 'clothes_hat', 'pillow', 'wall', 'oven', 'coffe_maker', 'television', 'curtain', 'diary', 'towel_rack', 'mouse', 'dishrack', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'microwave', 'cd_player', 'freezer', 'band-aids', 'bathroom', 'food_carrot', 'orchid', 'music_stand', 'phone', 'bookshelf', 'thread', 'mail', 'stereo', 'homework', 'mousepad', 'closetdrawer', 'clothes_hat', 'form', 'broom', 'newspaper', 'photoframe', 'check', 'food_rice', 'food_donut', 'dresser', 'sink', 'napkin', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to the living and turn to the music stand. Switch on the CD player to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <music_stand>', '[FIND] <music_stand>', '[TURNTO] <music_stand>', '[FIND] <cd_player>', '[SWITCHON] <cd_player>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'trashcan', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'thread', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'bench', 'cup', 'light_bulb', 'coffe_maker', 'toilet_paper', 'television', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "Type on a computer in the home office.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'food_ice_cream', 'tea', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'drawing', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'dining_room', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'cupboard', 'light', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the television. Switch on the television. Find the remote control. Grab the remote control. Point at television. Find the button. Push button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'bathroom', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'oven', 'check', 'television', 'oven_mitts', 'doorjamb', 'mouse', 'kitchen_counter', 'sink', 'napkin', 'table', 'shoe_rack']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I will walk to the computer at the desk in my home office, sit in the chair, turn on the computer, open a web browser, type in the address, and view the computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TOUCH] <mouse>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'bathroom', 'knifeblock', 'wallshelf', 'bathroom_counter', 'character', 'bedroom', 'tea_bag', 'keyboard', 'food_cake', 'toaster', 'ceilinglamp', 'shower', 'faucet', 'couch', 'food_jam', 'clothes_pants', 'bowl', 'dining_room', 'nightstand', 'oven', 'ice', 'lighter', 'mouse', 'conditioner', 'door', 'bathroom_cabinet', 'light']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to the living room, Find Phone stand. Find Phone. Pickup the receiver to attend the call.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['water_glass', 'tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'toaster', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'water', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I need to go to Kitchen.  Get a cup.  Fill cup.  Drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'tea', 'chair', 'drawing', 'orchid', 'cpuscreen', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'laptop', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'nightstand', 'oven', 'powersocket', 'bed', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to the home office. I put the laptop on the desk. I sit on the chair. I switch on my laptop and I type on the keyboard and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <laptop> <desk>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'glue', 'filing_cabinet', 'tablelamp', 'toilet', 'freezer', 'bathroom', 'cat', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'ceiling', 'bookshelf', 'keyboard', 'button', 'shoe-shine_kit', 'stereo', 'closetdrawer', 'nightstand', 'bench', 'cup', 'bed', 'photoframe', 'coffe_maker', 'curtain', 'shoes', 'dresser', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat.  Pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'orchid', 'bedroom', 'clothes_jacket', 'mop_bucket', 'stovefan', 'mousepad', 'pillow', 'dining_room', 'nightstand', 'wall', 'powersocket', 'oven', 'basket_for_clothes', 'bed', 'photoframe', 'mouthwash', 'television', 'wine', 'after_shave', 'doorjamb', 'sink', 'bathroom_cabinet', 'light', 'dishwasher', 'soap', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to the kitchen and find the basket of clothes. Put the soap and clothes into the washing machine. Turn on the washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <washing_machine>', '[TURNTO] <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[OPEN] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <clothes_jacket>', '[GRAB] <clothes_jacket>', '[PUTBACK] <clothes_jacket> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'bathroom', 'detergent', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'toilet_paper', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk towards to the bedroom, and I open the door of the bedroom, and I switch on the lights inside the bedroom, the switch is beside the door.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'phone', 'button', 'mat', 'stereo', 'window', 'homework', 'foundation', 'couch', 'pillow', 'wall', 'toy', 'cup', 'light_bulb', 'television', 'razor', 'curtain', 'after_shave', 'doorjamb', 'rag', 'coffee_pot', 'laundry_detergent', 'bathtub', 'table']\nTask:Arrange furniture\nresponse:\n{\n    \"thoughts\": Move furniture where you want it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[PUSH] <coffee_table>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'freezer', 'walllamp', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'toaster', 'shower', 'faucet', 'window', 'couch', 'wall', 'powersocket', 'bed', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub', 'light', 'laser_pointer']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "walk to bathroom , find shower , take a shower", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'detergent', 'cat', 'food_butter', 'knife', 'phone', 'character', 'table_cloth', 'mat', 'deck_of_cards', 'stereo', 'form', 'pillow', 'dining_room', 'nightstand', 'wall', 'basket_for_clothes', 'photoframe', 'tray', 'doorjamb', 'kitchen_counter', 'chessboard', 'dresser', 'clothes_dress', 'table', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['wallshelf', 'cpuscreen', 'hanger', 'floor', 'phone', 'desk', 'keyboard', 'bookshelf', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'window', 'dish_soap', 'form', 'nightstand', 'wall_clock', 'oven', 'powersocket', 'bench', 'tape', 'video_game_controller', 'doorjamb', 'kitchen_counter', 'dresser', 'home_office', 'light', 'dishwasher', 'bathroom_cabinet', 'table']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the dishwasher and put dish soap. I close the dishwasher and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'knife', 'desk', 'bookshelf', 'button', 'clothes_jacket', 'stovefan', 'deck_of_cards', 'mousepad', 'closetdrawer', 'oil', 'towel', 'pillow', 'broom', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'television', 'wine', 'curtain', 'after_shave', 'cards', 'home_office', 'light', 'clothes_skirt', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[WALK] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'hanger', 'bathroom_counter', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'plate', 'tvstand', 'faucet', 'shower', 'dish_soap', 'mousepad', 'wall', 'wall_clock', 'oven', 'photoframe', 'bed', 'television', 'tray', 'towel_rack', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk into kitchen, find dishwasher, put plates in dishwasher, put soap in dishwasher, turn on dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'pillow', 'wall', 'nightstand', 'bench', 'food_food', 'doorjamb', 'kettle', 'mouse', 'pencil', 'headset', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": go to the office, turn computer on, open an browser or any application to open any website.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'coffe_maker', 'television', 'doorjamb', 'centerpiece', 'mouse', 'cupboard', 'sink', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go find a computer, turn it on, click google, and browse the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'form', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'cupboard', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to livingroom, sit on couch, grab remote and turn on tv.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'knife', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'microphone', 'pillow', 'wall', 'powersocket', 'novel', 'television', 'doorjamb', 'towel_rack', 'mouse', 'chessboard', 'door', 'sink', 'home_office', 'table']\nTask:Work\nresponse:\n", "output": {"thoughts": "walk to home office ,find computer ,switch on computer and work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'crayon', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'ceilinglamp', 'clothes_socks', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'food_food', 'video_game_controller', 'doorjamb', 'diary', 'chessboard', 'dresser', 'door', 'light', 'table', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": flip the switch and turn on the lights,\n    \"actions\": ['[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'bathroom', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'stereo', 'ceilinglamp', 'closetdrawer', 'wall', 'nightstand', 'oven', 'newspaper', 'coffe_maker', 'mouse', 'dresser', 'door', 'bathroom_cabinet']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <address_book>", "[FIND] <address_book>", "[GRAB] <address_book>", "[READ] <address_book>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'food_salt', 'toilet', 'freezer', 'detergent', 'bills', 'orchid', 'hanger', 'ceiling', 'envelope', 'table_cloth', 'brush', 'mat', 'ceilinglamp', 'shower', 'stovefan', 'nightstand', 'powersocket', 'bench', 'photoframe', 'check', 'food_food', 'wooden_spoon', 'sink', 'napkin', 'instrument_guitar', 'table', 'coffee']\nTask:Pull up carpet\nresponse:\n{\n    \"thoughts\": Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mat>', '[FIND] <mat>', '[TOUCH] <mat>', '[PULL] <mat>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'character', 'bedroom', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'pasta', 'bench', 'cup', 'bed', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'headset', 'cupboard', 'instrument_guitar', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I bend down. I sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'bathroom', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'toaster', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'nightstand', 'television', 'toothbrush_holder', 'curtain', 'mouse', 'shoes', 'clothes_dress', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": go to bathroom. grab toothbrush. put toothpaste on toothbrush. brush teeth. turn on faucet. rinse toothbrush and mouth. put toothbrush away.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'chair', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'ceiling', 'desk', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'homework', 'mousepad', 'closetdrawer', 'pillow', 'dining_room', 'wall_clock', 'oven', 'tape', 'photoframe', 'dresser', 'cupboard', 'bathtub', 'light', 'dirt']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I reach out arm. I put hand on radio. I turn on radio. I pick a station.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[TOUCH] <stereo>", "[TURNTO] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_ice_cream', 'measuring_cup', 'knife', 'orchid', 'cpuscreen', 'bathroom_counter', 'ceiling', 'phone', 'alcohol', 'computer', 'table_cloth', 'mat', 'deck_of_cards', 'stereo', 'window', 'pillow', 'powersocket', 'cup', 'newspaper', 'check', 'ice', 'after_shave', 'lighter', 'diary', 'cd', 'cupboard', 'maindoor', 'colander', 'board_game']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find radio, turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'character', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'bench', 'tape', 'photoframe', 'television', 'curtain', 'cards', 'bathtub', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to living room and lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mop_bucket', 'mat', 'laptop', 'window', 'homework', 'mousepad', 'food_orange', 'dining_room', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'bed', 'tray', 'curtain', 'mouse', 'pencil', 'light']\nTask:Watch movie\nresponse:\n{\n    \"thoughts\": Bring laptop to bed and watch a movie,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[WALK] <bed>', '[SIT] <bed>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'trashcan', 'toothbrush', 'bathroom_counter', 'shampoo', 'floor', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'toaster', 'ceilinglamp', 'iron', 'shower', 'stereo', 'pillow', 'wall', 'wall_clock', 'photoframe', 'food_food', 'water', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[FIND] <toothbrush>", "[FIND] <water>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_ice_cream', 'chair', 'nail_polish', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'ceilinglamp', 'balanceball', 'shower', 'mousepad', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'toothbrush_holder', 'doorjamb', 'mouse', 'light', 'vase']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'detergent', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'shoe-shine_kit', 'faucet', 'clothes_socks', 'closetdrawer', 'mousepad', 'dining_room', 'broom', 'wall', 'nightstand', 'photoframe', 'television', 'water', 'tray', 'cards', 'paper_towel', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. I turn on the water. I put soap on my hands. I rinse my hands. I turn off the water. I dry my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <water>", "[TURNTO] <water>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[TURNTO] <water>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'bathroom', 'detergent', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_peanut_butter', 'pillow', 'towel', 'dining_room', 'wall', 'powersocket', 'toilet_paper', 'television', 'curtain', 'pencil', 'headset', 'cd', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Lift the remote and change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'drawing', 'orchid', 'cpuscreen', 'phone', 'desk', 'bedroom', 'bookshelf', 'thread', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'closetdrawer', 'pillow', 'nightstand', 'wall_clock', 'bed', 'light_bulb', 'photoframe', 'tray', 'doorjamb', 'mouse', 'cupboard', 'door', 'bathtub', 'sink', 'light', 'clothes_shirt']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "go to the bathroom and brush your teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'cpuscreen', 'phone', 'desk', 'cleaning_solution', 'keyboard', 'purse', 'table_cloth', 'stovefan', 'tvstand', 'shower', 'stereo', 'food_noodles', 'towel', 'pillow', 'wall', 'wall_clock', 'oven', 'basket_for_clothes', 'photoframe', 'novel', 'food_rice', 'rag', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'napkin', 'table']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'trashcan', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'ceilinglamp', 'mat', 'faucet', 'shower', 'couch', 'towel', 'wall', 'bench', 'oven', 'photoframe', 'bed', 'mouse', 'kitchen_counter', 'dresser', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'detergent', 'food_butter', 'dough', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'purse', 'faucet', 'stereo', 'picture', 'dustpan', 'bowl', 'dining_room', 'wall', 'oven', 'powersocket', 'photoframe', 'wine', 'tray', 'console', 'pencil', 'sink', 'bathroom_cabinet', 'dishwasher', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'chair', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'toothbrush', 'needle', 'desk', 'keyboard', 'alcohol', 'tvstand', 'faucet', 'window', 'food_noodles', 'couch', 'form', 'pillow', 'wall', 'wall_clock', 'bench', 'bed', 'curtain', 'cupboard', 'instrument_violin', 'bathtub', 'home_office', 'light', 'board_game']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "i go find the phone. i pick up the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sponge', 'food_bread', 'blender', 'chair', 'walllamp', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'tvstand', 'window', 'mousepad', 'clothes_socks', 'wall', 'oven', 'cup', 'television', 'mouse', 'cupboard', 'door', 'home_office', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": go to the office, turn computer on, open an browser or any application to open any website.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <computer>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'microwave', 'toilet', 'chair', 'freezer', 'band-aids', 'trashcan', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'character', 'food_bacon', 'mat', 'tvstand', 'shower', 'pillow', 'wall', 'bed', 'photoframe', 'food_food', 'doorjamb', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'door']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'beer', 'bathroom', 'detergent', 'trashcan', 'cpuscreen', 'ceiling', 'bedroom', 'desk', 'keyboard', 'clothes_jacket', 'tvstand', 'couch', 'pillow', 'food_kiwi', 'oven', 'bench', 'coffe_maker', 'oven_mitts', 'towel_rack', 'pencil', 'cupboard', 'hair', 'clothes_dress', 'food_steak', 'napkin', 'vase', 'laser_pointer']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into the bedroom. I walk to stereo. I turn on stereo.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'mousepad', 'wall', 'nightstand', 'powersocket', 'coffe_maker', 'check', 'television', 'wine', 'doorjamb', 'mouse', 'cupboard', 'controller', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to my desk. I turn on the computer. I open a web page and look at it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'laptop', 'faucet', 'mousepad', 'candle', 'microphone', 'food_orange', 'wall', 'photoframe', 'check', 'mouthwash', 'after_shave', 'doorjamb', 'cupboard', 'pajamas', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'light', 'soap', 'clothes_shirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go into the bathroom and turn on the light. I walk to the sink and turn on the water. I grab the soap and wash my hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[PUTOBJBACK] <soap>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'mousepad', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to the Home Office.  Find the computer.  Turn on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'food_carrot', 'drawing', 'hanger', 'phone', 'ceiling', 'desk', 'computer', 'toaster', 'mat', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'towel', 'cup', 'food_cheese', 'keys', 'tray', 'after_shave', 'towel_rack', 'mouse', 'cupboard', 'sink', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'trashcan', 'orchid', 'floor', 'phone', 'character', 'keyboard', 'envelope', 'clothes_jacket', 'piano_bench', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'wall', 'nightstand', 'bench', 'wall_clock', 'television', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[TURNTO] <toilet>", "[LOOKAT] <toilet>", "[WALK] <toilet>", "[TURNTO] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'sheets', 'tablelamp', 'toilet', 'freezer', 'bathroom', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'pillow', 'oven', 'bed', 'coffe_maker', 'doorjamb', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'mat', 'shower', 'mousepad', 'wall', 'powersocket', 'oven', 'bench', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'door', 'table', 'light', 'coffee']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go in my office, sit at my desk turn on the computer, then type on my keyboard and open my web browser.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'knife', 'knifeblock', 'orchid', 'drawing', 'floor', 'phone', 'character', 'bedroom', 'remote_control', 'computer', 'button', 'mat', 'tvstand', 'faucet', 'shower', 'oil', 'dining_room', 'powersocket', 'newspaper', 'ice', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'door', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I pick up the remote. I press button on remote.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'character', 'food_bacon', 'mat', 'ceilinglamp', 'shower', 'wall', 'wall_clock', 'cup', 'photoframe', 'bed', 'food_food', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'coffee']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[TURNTO] <cup>", "[LOOKAT] <cup>", "[WALK] <cup>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[WALK] <freezer>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'faucet', 'picture', 'window', 'dustpan', 'towel', 'wall', 'nightstand', 'photoframe', 'keys', 'toothbrush_holder', 'food_donut', 'mouse', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Cleaning hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'trashcan', 'drawing', 'orchid', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'computer', 'toaster', 'deck_of_cards', 'mousepad', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'powersocket', 'bench', 'television', 'doorjamb', 'cd', 'dresser', 'conditioner', 'door', 'home_office', 'light', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the phone. I pick up the phone. I hold it to my ear.", "actions": ["[WALK] <dining_room>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'walllamp', 'ground_coffee', 'orchid', 'drawing', 'cpuscreen', 'floor', 'phone', 'ceiling', 'toaster', 'ceilinglamp', 'plate', 'tvstand', 'window', 'dish_soap', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'coffe_maker', 'razor', 'tray', 'mouse', 'kitchen_counter', 'cupboard', 'bathtub', 'dishwasher', 'instrument_violin']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['chair', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'stovefan', 'toaster', 'laptop', 'faucet', 'closetdrawer', 'couch', 'bowl', 'dining_room', 'wall', 'oven', 'bench', 'bed', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'bathtub']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I switch on the stove and cook my groceries there.", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[SWITCHON] <oven>", "[FIND] <bowl>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <oven>", "[FIND] <food_food>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'drawing', 'knifeblock', 'floor', 'ceiling', 'ceilinglamp', 'mat', 'shower', 'picture', 'window', 'food_noodles', 'food_jam', 'wall', 'powersocket', 'oven', 'photoframe', 'check', 'television', 'toothbrush_holder', 'water', 'curtain', 'doorjamb', 'kitchen_counter', 'cd', 'headset', 'door', 'bathroom_cabinet', 'light', 'soap']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'novel', 'television', 'tray', 'curtain', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'home_office', 'table', 'bathroom_cabinet', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Find the sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'band-aids', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'mousepad', 'oil', 'broom', 'wall', 'nightstand', 'bench', 'newspaper', 'coffe_maker', 'food_food', 'coin', 'food_donut', 'lighter', 'mouse', 'chessboard', 'door', 'table', 'dirt']\nTask:Turking\nresponse:\n{\n    \"thoughts\": Do a random variety very small jobs from different clients on mturk.Com,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'keyboard', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'food_kiwi', 'wall', 'light_bulb', 'novel', 'television', 'curtain', 'cupboard', 'bathtub', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I go to the living room, take a book from the shelf, sit in the sofa and read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'crayon', 'mop', 'toilet', 'chair', 'bathroom', 'detergent', 'dough', 'ceiling', 'keyboard', 'stereo', 'couch', 'pillow', 'cup', 'newspaper', 'keys', 'check', 'toothbrush_holder', 'food_donut', 'after_shave', 'rag', 'mouse', 'pajamas', 'cupboard', 'bag', 'light', 'table', 'tooth_paste']\nTask:Put away keys\nresponse:\n{\n    \"thoughts\": Return keys into a bag,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <keys>', '[FIND] <keys>', '[GRAB] <keys>', '[FIND] <bag>', '[OPEN] <bag>', '[PUTBACK] <keys> <bag>', '[CLOSE] <bag>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'window', 'electric_shaver', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "watching TV from sofa", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[DROP] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'drawing', 'orchid', 'wallshelf', 'floor', 'desk', 'bookshelf', 'purse', 'computer', 'mat', 'tvstand', 'couch', 'clothes_pants', 'pillow', 'wall', 'bench', 'powersocket', 'basket_for_clothes', 'photoframe', 'bed', 'television', 'doorjamb', 'pencil', 'laundry_detergent', 'clothes_dress', 'table', 'light', 'clothes_shirt', 'washing_machine']\nTask:Do laundry\nresponse:\n{\n    \"thoughts\": Collecting laundry from the children's hamper,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[PUTBACK] <basket_for_clothes> <table>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[POUR] <laundry_detergent> <washing_machine>', '[PUTOBJBACK] <laundry_detergent>', '[CLOSE] <filing_cabinet>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[PLUGIN] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'trashcan', 'drawing', 'orchid', 'ceiling', 'bedroom', 'mail', 'mat', 'tvstand', 'mousepad', 'dining_room', 'wall', 'wall_clock', 'basket_for_clothes', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'door', 'clothes_dress', 'bathroom_cabinet', 'table', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'bed', 'television', 'doorjamb', 'shaving_cream', 'mouse', 'coffee', 'dresser', 'hairbrush', 'clothes_scarf', 'soap', 'teeth']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sitting on the couch watching my favorite television show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'ground_coffee', 'walllamp', 'detergent', 'trashcan', 'drawing', 'wallshelf', 'character', 'remote_control', 'button', 'shower', 'faucet', 'window', 'pillow', 'wall', 'bench', 'newspaper', 'coffee_table', 'photoframe', 'coffe_maker', 'television', 'food_donut', 'tray', 'sink', 'napkin', 'bathroom_cabinet', 'home_office', 'instrument_guitar', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'check', 'television', 'doorjamb', 'mouse', 'address_book', 'dresser', 'door']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Walk to Home office. Sit on the chair. Switch on the computer. Find address in addressbook. Type in the keyboard. Write mail.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <address_book>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'stamp', 'bedroom', 'stovefan', 'toaster', 'plate', 'faucet', 'shower', 'dish_soap', 'food_jam', 'towel', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'ice', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'chair', 'bathroom', 'orchid', 'cpuscreen', 'bathroom_counter', 'desk', 'keyboard', 'button', 'piano_bench', 'ceilinglamp', 'stereo', 'floor_lamp', 'oil', 'mousepad', 'closetdrawer', 'tape', 'love_seat', 'newspaper', 'bed', 'check', 'video_game_controller', 'mouthwash', 'curtain', 'lighter', 'diary', 'doorjamb', 'dresser', 'sink']\nTask:Read\nresponse:\n{\n    \"thoughts\": Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'chair', 'freezer', 'bathroom', 'trashcan', 'orchid', 'cpuscreen', 'hanger', 'character', 'computer', 'mail', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'bed', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'cupboard', 'sink', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the cat. I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'tablelamp', 'food_salt', 'microwave', 'chair', 'freezer', 'ground_coffee', 'dough', 'drawing', 'orchid', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'toaster', 'mat', 'ceilinglamp', 'stovefan', 'faucet', 'shower', 'oil', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'powersocket', 'food_food', 'coffe_maker', 'television', 'wine', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'food_steak', 'door', 'sink', 'table', 'dishwasher', 'light']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Walk into kitchen. Open fridge. Take food from fridge. Close fridge. Put food on kitchen counter. Prepare food. Turn on stove. Open kitchen cabinet. Take pot. Close cabinet. Put food in pot. Put pot on stove. Cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <kitchen_counter>', '[PUTBACK] <food_food> <kitchen_counter>', '[CLOSE] <freezer>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pot>', '[GRAB] <pot>', '[CLOSE] <kitchen_cabinet>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <oven>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'freezer', 'detergent', 'floor', 'phone', 'ceiling', 'bedroom', 'keyboard', 'stovefan', 'shower', 'faucet', 'window', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'bench', 'photoframe', 'coffe_maker', 'check', 'television', 'food_food', 'ice', 'rag', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go home and it's dark so I can not see anything", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'freezer', 'walllamp', 'food_butter', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'alcohol', 'window', 'food_jam', 'dining_room', 'wall', 'bench', 'powersocket', 'newspaper', 'food_food', 'doorjamb', 'kettle', 'towel_rack', 'dresser', 'cupboard', 'door', 'table', 'light']\nTask:Eat dinner\nresponse:\n{\n    \"thoughts\": In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'keys', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I want to relax on the sofa to some relaxing music in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <stereo>", "[PLUGIN] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'tea', 'beer', 'microwave', 'toilet', 'trashcan', 'food_carrot', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'bedroom', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'window', 'dining_room', 'food_kiwi', 'nightstand', 'bed', 'photoframe', 'doorjamb', 'dishrack', 'cupboard', 'sink', 'home_office']\nTask:Open bathroom window\nresponse:\n{\n    \"thoughts\": Make sure bathroom window is unlocked. Lift up bathroom window so it is open.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <window>', '[FIND] <window>', '[OPEN] <window>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'microwave', 'toilet', 'walllamp', 'drawing', 'needle', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'desk', 'computer', 'bookmark', 'stovefan', 'plate', 'tvstand', 'dish_soap', 'closetdrawer', 'dining_room', 'wall', 'wall_clock', 'oven', 'powersocket', 'curtain', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_salt', 'chair', 'detergent', 'trashcan', 'dough', 'drawing', 'character', 'computer', 'button', 'brush', 'mat', 'couch', 'clothes_pants', 'clothes_hat', 'wall', 'nightstand', 'basket_for_clothes', 'food_cheese', 'rag', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'laundry_detergent', 'conditioner', 'sink', 'home_office', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['spectacles', 'microwave', 'freezer', 'walllamp', 'trashcan', 'floor', 'desk', 'keyboard', 'clothes_jacket', 'piano_bench', 'tvstand', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'food_kiwi', 'nightstand', 'oven', 'bench', 'powersocket', 'basket_for_clothes', 'photoframe', 'tray', 'dresser', 'cupboard', 'home_office', 'light', 'clothes_skirt', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and find the basket of clothes. Put the soap and clothes into the washing machine. Turn on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <washing_machine>", "[TURNTO] <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'toilet', 'dog', 'detergent', 'orchid', 'wallshelf', 'food_egg', 'cpuscreen', 'phone', 'stamp', 'bedroom', 'bookshelf', 'bookmark', 'shower', 'mousepad', 'food_jam', 'towel', 'pasta', 'bed', 'novel', 'food_food', 'curtain', 'centerpiece', 'shaving_cream', 'dresser', 'cupboard', 'table', 'creditcard']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the entrance hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'chair', 'bathroom_counter', 'floor', 'laptop', 'faucet', 'shower', 'towel', 'wall', 'wall_clock', 'nightstand', 'cup', 'bed', 'photoframe', 'water', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go into the bathroom and go over to the sink. Grab soap and turn on the water. Wet your hands and rub them together so that the soap lathers. Rub the soap all over your hands thoroughly. Rinse off the soap with water and turn off the sink. Dry your hands with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'glue', 'toilet', 'chair', 'detergent', 'dough', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'keyboard', 'alcohol', 'stovefan', 'ceilinglamp', 'couch', 'microphone', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'doorjamb', 'kitchen_counter', 'hairbrush', 'cupboard', 'home_office', 'dishwasher', 'coffee', 'shoe_rack']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'bathroom', 'hanger', 'floor', 'phone', 'ceiling', 'character', 'bedroom', 'bookshelf', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'closetdrawer', 'couch', 'pillow', 'cup', 'coffe_maker', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'home_office', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'thread', 'computer', 'piano_bench', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'photoframe', 'bed', 'television', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to the home office and turn to the desk. Switch on the computer. Grab the mouse and keyboard to start working.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[TURNTO] <desk>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <mouse>', '[GRAB] <mouse>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'trashcan', 'drawing', 'knife', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'wall_clock', 'powersocket', 'oven', 'photoframe', 'check', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'controller', 'bathtub', 'home_office', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find cat. Walk to cat. Pull cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[PULL] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'cup', 'newspaper', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Start computer\nresponse:\n{\n    \"thoughts\": Sit in desk chair, press power button, let computer power on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'floor', 'alcohol', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'deck_of_cards', 'window', 'shower', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'photoframe', 'keys', 'television', 'curtain', 'doorjamb', 'mouse', 'home_office', 'table', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into bedrrom and to the sofa. Sit on the sofa and touch cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'shampoo', 'character', 'wine_glass', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'couch', 'pillow', 'wall', 'bench', 'powersocket', 'food_cheese', 'food_food', 'coffe_maker', 'television', 'wine', 'curtain', 'table']\nTask:Entertain\nresponse:\n{\n    \"thoughts\": Give guests wine, cheese, crackers and sit and talk,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <food_cheese>', '[FIND] <food_cheese>', '[FIND] <food_food>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <wine_glass>', '[GRAB] <wine_glass>', '[DRINK] <wine_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'measuring_cup', 'chair', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'dining_room', 'wall', 'nightstand', 'doorjamb', 'mouse', 'dresser', 'controller', 'laundry_detergent', 'door', 'light', 'table', 'scrabble', 'dirt']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk to the bedroom, then find the computer and sit in the chair ,next switch on the computer and finally type in the keyboard to start to work on it", "actions": ["[WALK] <bedroom>", "[WALK] <computer>", "[FIND] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'glue', 'toilet', 'chair', 'detergent', 'dough', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'keyboard', 'alcohol', 'stovefan', 'ceilinglamp', 'couch', 'microphone', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'coffee_filter', 'doorjamb', 'kitchen_counter', 'hairbrush', 'cupboard', 'home_office', 'dishwasher', 'coffee', 'shoe_rack']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'tea_bag', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'bench', 'photoframe', 'television', 'tray', 'doorjamb', 'mouse', 'home_office', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to home office. Find computer. Sit in chair in front of computer desk. Turn on computer. Open internet browser. Browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'bills', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'television', 'curtain', 'doorjamb', 'door', 'table', 'light']\nTask:Watch  TV\nresponse:\n{\n    \"thoughts\": Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'dog', 'freezer', 'orchid', 'drawing', 'wallshelf', 'ceiling', 'character', 'mail', 'clothes_jacket', 'computer', 'mat', 'faucet', 'window', 'closetdrawer', 'nightstand', 'wall_clock', 'oven', 'basket_for_clothes', 'coffe_maker', 'mouse', 'dresser', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'light', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Pick up my dirty clothes, load the washer, add detergent, turn on washer", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[RUN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Using desktop computer in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'toothbrush', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'table_cloth', 'stovefan', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'towel', 'wall', 'bed', 'coffe_maker', 'door', 'sink', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I go to sleep and i eat the dinner so i have to clean my mouth", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[DROP] <tooth_paste>", "[FIND] <teeth>", "[SCRUB] <teeth>", "[RINSE] <teeth>", "[FIND] <towel>", "[GRAB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'shaving_cream', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": switch on the computer,  grab a chair and sit in front of the computer. type with keyboard and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'detergent', 'trashcan', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'closetdrawer', 'wall', 'oven', 'cup', 'novel', 'water', 'drinking_glass', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'door']\nTask:Drink\nresponse:\n", "output": {"thoughts": "go to the kitchen", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <water>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'homework', 'towel', 'dining_room', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bag', 'napkin', 'soap']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen counter, grab a grocery, walk to fridge, open fridge, put the grocery into fridge, and close fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'freezer', 'walllamp', 'detergent', 'cat', 'knife', 'drawing', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'phone', 'ceiling', 'character', 'faucet', 'window', 'foundation', 'closetdrawer', 'food_orange', 'dining_room', 'coffe_maker', 'check', 'television', 'food_food', 'tray', 'curtain', 'mouse', 'pajamas', 'door', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room and I pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'tablelamp', 'microwave', 'dog', 'milk', 'orchid', 'floor', 'ceiling', 'computer', 'toaster', 'plate', 'shower', 'dish_soap', 'wall', 'food_cheese', 'coffe_maker', 'check', 'coffee_filter', 'oven_mitts', 'cards', 'towel_rack', 'pencil', 'kitchen_counter', 'comb', 'cupboard', 'conditioner', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the  kitchen. Open the dishwasher. Put plates into the dishwasher. Put dish soap into the dishwasher. Close the dishwasher. Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'ground_coffee', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'stovefan', 'toaster', 'tvstand', 'faucet', 'shower', 'closetdrawer', 'wall', 'nightstand', 'oven', 'bench', 'tape', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'diary', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'toilet', 'chair', 'bathroom', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'computer', 'tvstand', 'faucet', 'stereo', 'food_orange', 'photoframe', 'novel', 'television', 'curtain', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'shoes', 'dresser', 'cupboard', 'sink', 'bathtub', 'home_office']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": I get dressed in my bedroom in the mornings.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'bathroom', 'trashcan', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'keyboard', 'bookshelf', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'window', 'food_peanut_butter', 'clothes_pants', 'oven', 'basket_for_clothes', 'food_food', 'television', 'kettle', 'dresser', 'laundry_detergent', 'door', 'bathtub', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'dog', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'laptop', 'mat', 'mousepad', 'wall', 'oven', 'love_seat', 'bed', 'coffe_maker', 'food_food', 'doorjamb', 'centerpiece', 'coffee_pot', 'mouse', 'pencil', 'hairbrush', 'controller', 'door', 'dishwasher', 'electrical_outlet']\nTask:Write book\nresponse:\n{\n    \"thoughts\": Writing a book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <electrical_outlet>', '[PLUGIN] <laptop>', '[WALK] <love_seat>', '[SIT] <love_seat>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'computer', 'food_bacon', 'mat', 'tvstand', 'ceilinglamp', 'plate', 'window', 'shower', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'wine', 'curtain', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'food_carrot', 'drawing', 'orchid', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'computer', 'mat', 'plate', 'ceilinglamp', 'stereo', 'couch', 'candle', 'food_orange', 'pillow', 'wall', 'oven', 'food_vegetable', 'novel', 'food_food', 'razor', 'television', 'console', 'kitchen_counter', 'cupboard', 'controller', 'table']\nTask:Fix snack\nresponse:\n{\n    \"thoughts\": Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTBACK] <food_carrot> <plate>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[PUTBACK] <food_vegetable> <plate>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'ground_coffee', 'drawing', 'orchid', 'cpuscreen', 'phone', 'bedroom', 'stamp', 'desk', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'nightstand', 'wall', 'oven', 'wall_clock', 'photoframe', 'coffe_maker', 'mouse', 'dresser', 'cupboard', 'bathtub', 'sink', 'home_office', 'table', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "You walk to the living room and find the cat. You look at the cat and touch it to pet it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_ice_cream', 'beer', 'microwave', 'freezer', 'walllamp', 'floor', 'ceiling', 'phone', 'dvd_player', 'stovefan', 'ceilinglamp', 'toaster', 'mousepad', 'wall', 'wall_clock', 'cup', 'bed', 'food_food', 'coin', 'food_rice', 'water', 'pencil', 'shoes', 'controller', 'napkin', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'bookmark', 'stovefan', 'mat', 'window', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'coffe_maker', 'curtain', 'doorjamb', 'door', 'sink', 'light', 'laser_pointer']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I go to the bedroom and turn on the light. I lie in bed. I grab my book from the nightstand, and look for the bookmark to check where I last left off. I read for a while. When I'm done, I put the bookmark on the page I stop at and place the book back on the nightstand.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[READ] <novel>", "[PUTBACK] <bookmark> <novel>", "[FIND] <nightstand>", "[PUTBACK] <novel> <nightstand>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'microwave', 'chair', 'food_butter', 'drawing', 'bathroom_counter', 'floor', 'stamp', 'desk', 'bookmark', 'stovefan', 'tvstand', 'shower', 'microphone', 'towel', 'dining_room', 'bowl', 'nightstand', 'oven', 'bench', 'shelf', 'food_cheese', 'photoframe', 'light_bulb', 'tray', 'kitchen_counter', 'dresser', 'hair', 'door', 'bathtub']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Go into the living room and pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>', '[TOUCH] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'freezer', 'bathroom', 'needle', 'cpuscreen', 'hanger', 'bathroom_counter', 'phone', 'character', 'desk', 'bedroom', 'bookshelf', 'mat', 'ceilinglamp', 'faucet', 'couch', 'pillow', 'wall_clock', 'basket_for_clothes', 'cup', 'bed', 'tray', 'kitchen_counter', 'cupboard', 'door', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "put clothes and soap in the washing machine, turn it on", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'food_food', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink']\nTask:Work\nresponse:\n{\n    \"thoughts\": working on computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <button>', '[TURNTO] <button>', '[LOOKAT] <button>', '[FIND] <mouse>', '[GRAB] <mouse>', '[GRAB] <button>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'television', 'tray', 'curtain', 'dresser', 'bathtub', 'sink', 'bathroom_cabinet', 'coffee_cup', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'blender', 'microwave', 'chair', 'walllamp', 'cpuscreen', 'bathroom_counter', 'ceiling', 'character', 'mail', 'food_bacon', 'mat', 'stereo', 'faucet', 'homework', 'shower', 'dining_room', 'wall', 'oven', 'keys', 'television', 'food_onion', 'coffee_pot', 'towel_rack', 'kitchen_counter', 'laundry_detergent', 'bag', 'spoon', 'table']\nTask:Put away keys\nresponse:\n{\n    \"thoughts\": Return keys into a bag,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <keys>', '[FIND] <keys>', '[GRAB] <keys>', '[FIND] <bag>', '[OPEN] <bag>', '[PUTBACK] <keys> <bag>', '[CLOSE] <bag>']\n}\n", "input": "\nScene_Objects:['sheets', 'freezer', 'walllamp', 'trashcan', 'drawing', 'orchid', 'ceiling', 'phone', 'bedroom', 'stovefan', 'mat', 'tvstand', 'shower', 'faucet', 'iron', 'food_noodles', 'couch', 'towel', 'pillow', 'dining_room', 'wall_clock', 'bench', 'bed', 'television', 'after_shave', 'doorjamb', 'door', 'bathtub', 'spoon', 'bathroom_cabinet']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to home office 1. Walk over to phone 1 on table 1.  Grab phone 1.  lift phone 1.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'scissors', 'juice', 'food_salt', 'microwave', 'chair', 'food_butter', 'toothbrush', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'button', 'brush', 'ceilinglamp', 'plate', 'wall', 'tape', 'food_food', 'check', 'food_donut', 'wooden_spoon', 'rag', 'door', 'clothes_scarf', 'napkin', 'table', 'light', 'dishwasher']\nTask:Eat\nresponse:\n{\n    \"thoughts\": Serve food on dinning table in appropriate utensils, sit on the chair, eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <plate>', '[PUTBACK] <food_food> <plate>', '[GRAB] <food_food>', '[PUTOBJBACK] <food_food>', '[FIND] <fork>', '[GRAB] <fork>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['blender', 'microwave', 'chair', 'freezer', 'bathroom', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'stereo', 'picture', 'mat', 'closetdrawer', 'couch', 'dining_room', 'wall_clock', 'bench', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'home_office', 'instrument_guitar']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the living room and then I walk to the light switch in there and turn the lights on", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'chair', 'walllamp', 'detergent', 'food_carrot', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'table_cloth', 'mop_bucket', 'ceilinglamp', 'tvstand', 'mat', 'couch', 'dustpan', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'check', 'television', 'tray', 'wooden_spoon', 'kettle', 'controller', 'food_steak', 'sink']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room and sit on the couch, ease into the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'shower', 'dish_soap', 'mousepad', 'pillow', 'wall', 'oven', 'photoframe', 'bed', 'coffe_maker', 'curtain', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'light']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "i go to the kitchen, i find the faucet. I turn the faucet on, i grab the plate to wash , i then get the soap. I wash the dishes by hand", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'blender', 'walllamp', 'orchid', 'cpuscreen', 'hanger', 'floor', 'bedroom', 'bookshelf', 'stovefan', 'food_peanut_butter', 'clothes_hat', 'clothes_pants', 'dining_room', 'nightstand', 'wall_clock', 'video_game_controller', 'food_donut', 'curtain', 'kettle', 'shoes', 'dresser', 'cupboard', 'hairbrush', 'controller', 'home_office', 'light', 'clothes_skirt', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the radio. Turn on the radio. Listen to the radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'walllamp', 'drawing', 'orchid', 'floor', 'phone', 'character', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'wall_clock', 'bed', 'television', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'hairdryer', 'sink', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the sofa. Sit down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'stereo', 'window', 'shower', 'mousepad', 'wall', 'powersocket', 'bench', 'clothes_gloves', 'television', 'tray', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'bathtub', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk into home office. I walk to desk. I sit at chair. I turn on computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sponge', 'sheets', 'tablelamp', 'toilet', 'freezer', 'bathroom', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'pillow', 'oven', 'bed', 'coffe_maker', 'doorjamb', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'home_office']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'chair', 'milk', 'toothbrush', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'alcohol', 'mop_bucket', 'table_cloth', 'mat', 'faucet', 'homework', 'clothes_socks', 'closetdrawer', 'teeth', 'mousepad', 'wall', 'powersocket', 'doorjamb', 'shaving_cream', 'sink', 'bathroom_cabinet', 'light', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'form', 'wall', 'powersocket', 'photoframe', 'video_game_controller', 'television', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'light', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk in to the bathroom. I find the soap. I find the faucet. I turn on the faucet. I wash my hands. I put the soap back. I rinse my hands. I turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'mop', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'envelope', 'faucet', 'picture', 'homework', 'window', 'mousepad', 'clothes_hat', 'towel', 'wall', 'photoframe', 'hands_both', 'food_food', 'doorjamb', 'kettle', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'light', 'vase', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I need to rinse my dentures off and put them back in my mouth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <teeth>', '[GRAB] <teeth>', '[WASH] <teeth>', '[PUTOBJBACK] <teeth>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <towel>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'check', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'sink', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to the office and sit in a chair, I turn on the computer and grab the mouse. I type on the keyboard and starting working on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'crayon', 'microwave', 'detergent', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'alcohol', 'toaster', 'shower', 'couch', 'clothes_hat', 'pillow', 'wall', 'pasta', 'bench', 'oven', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'doorjamb', 'hairbrush', 'cupboard', 'door', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'trashcan', 'toothbrush', 'drawing', 'needle', 'floor', 'ceiling', 'phone', 'computer', 'button', 'piano_bench', 'plate', 'faucet', 'pillow', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'bed', 'food_food', 'television', 'doorjamb', 'mouse', 'kitchen_counter', 'conditioner', 'door', 'food_steak', 'bathtub', 'light', 'teeth']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch. Turn on lights.", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'drawing', 'keyboard', 'bookmark', 'stovefan', 'ceilinglamp', 'stereo', 'shower', 'foundation', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'bed', 'check', 'television', 'wine', 'curtain', 'towel_rack', 'vacuum_cleaner', 'maindoor', 'door', 'dishwasher', 'vase']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter living room. Walk to telephone. Pick up from hook.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'drawing', 'wallshelf', 'food_egg', 'hanger', 'cpuscreen', 'floor', 'ceiling', 'character', 'keyboard', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'toaster', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'bed', 'television', 'curtain', 'doorjamb', 'dresser', 'light', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and turn on the light. I walk to the couch then sit on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'sheets', 'pot', 'blender', 'microwave', 'chair', 'nail_polish', 'detergent', 'milk', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'clothes_socks', 'broom', 'wall', 'food_kiwi', 'powersocket', 'cup', 'photoframe', 'keys', 'newspaper', 'curtain', 'lighter', 'pencil', 'pajamas', 'clothes_dress', 'sink', 'table', 'light']\nTask:Sit in chair\nresponse:\n{\n    \"thoughts\": I sit in a chair to rest and eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['chair', 'band-aids', 'walllamp', 'bathroom', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'pillow', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'bed', 'food_food', 'check', 'television', 'doorjamb', 'mouse', 'sink', 'napkin', 'table', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "go to the home office and write an email", "actions": ["[RUN] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <mouse>", "[GRAB] <mouse>", "[TOUCH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[TOUCH] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[TYPE] <keyboard>", "[TOUCH] <mouse>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'crayon', 'freezer', 'bathroom', 'toothbrush', 'drawing', 'knifeblock', 'floor', 'phone', 'mail', 'brush', 'ceilinglamp', 'picture', 'window', 'couch', 'bowl', 'wall', 'pasta', 'bench', 'tape', 'newspaper', 'light_bulb', 'television', 'toothbrush_holder', 'shaving_cream', 'vacuum_cleaner', 'spoon', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I get home from work and walk into the living room. I see the couch and go to sit on it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'bathroom', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'photoframe', 'television', 'coffee_filter', 'curtain', 'towel_rack', 'dresser', 'cupboard', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'computer', 'stovefan', 'faucet', 'window', 'mousepad', 'closetdrawer', 'couch', 'towel', 'wall', 'nightstand', 'curtain', 'towel_rack', 'mouse', 'coffee_pot', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom and locate the sink. I turn the faucet so that water is running, and rinse my hands briefly. Then I grab the soap and squeeze some onto my hands. I spend a few minutes rubbing my hands together, then I rinse them again. Finally, I turn the faucet off and wipe my hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SQUEEZE] <soap>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'keyboard', 'bookshelf', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'bench', 'bed', 'check', 'television', 'tray', 'curtain', 'mouse', 'cupboard', 'door', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'scissors', 'chair', 'drawing', 'wallshelf', 'floor', 'bookshelf', 'envelope', 'piano_bench', 'toaster', 'ceilinglamp', 'plate', 'shower', 'picture', 'oil', 'closetdrawer', 'microphone', 'food_cheese', 'newspaper', 'bed', 'food_food', 'wine', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'light']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.,\n    \"actions\": ['[WALK] <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'toaster', 'mat', 'ceilinglamp', 'shower', 'faucet', 'food_jam', 'dining_room', 'nightstand', 'wall', 'bench', 'powersocket', 'video_game_controller', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'cat', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'clothes_jacket', 'stovefan', 'ceilinglamp', 'shower', 'mousepad', 'couch', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'box', 'food_food', 'television', 'coffee_pot', 'kettle', 'cards', 'mouse', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'water_glass', 'microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'wallshelf', 'bookshelf', 'clothes_jacket', 'toaster', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'dining_room', 'nightstand', 'wall', 'wall_clock', 'powersocket', 'basket_for_clothes', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'console', 'dishrack', 'hairdryer', 'door', 'bathtub']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drink water", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'bathroom', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bookshelf', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'couch', 'dining_room', 'food_kiwi', 'wall', 'toilet_paper', 'razor', 'television', 'curtain', 'diary', 'towel_rack', 'door', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch and use the remote control to turn on the TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'bathroom', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'keyboard', 'computer', 'stovefan', 'faucet', 'window', 'mousepad', 'closetdrawer', 'couch', 'towel', 'wall', 'nightstand', 'curtain', 'towel_rack', 'mouse', 'coffee_pot', 'dresser', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom and locate the sink. I turn the faucet so that water is running, and rinse my hands briefly. Then I grab the soap and squeeze some onto my hands. I spend a few minutes rubbing my hands together, then I rinse them again. Finally, I turn the faucet off and wipe my hands on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SQUEEZE] <soap>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'dog', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'folder', 'desk', 'bookshelf', 'keyboard', 'thread', 'computer', 'mail', 'bookmark', 'mousepad', 'wall', 'nightstand', 'photoframe', 'novel', 'toilet_paper', 'tray', 'towel_rack', 'mouse', 'door', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find a desk and grab a book. Walk to the chair, sit in the chair and read a book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'hanger', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'shower', 'closetdrawer', 'microphone', 'wall', 'nightstand', 'oven', 'wall_clock', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'door', 'bathroom_cabinet', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "flip the switch and turn on the lights", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'freezer', 'food_carrot', 'milk', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'mousepad', 'garbage_can', 'clothes_hat', 'food_orange', 'food_dessert', 'wall', 'powersocket', 'food_vegetable', 'food_cheese', 'doorjamb', 'lighter', 'mouse', 'coffee', 'door', 'light', 'colander', 'teeth']\nTask:Rotate stock in refrigerator\nresponse:\n{\n    \"thoughts\": Throw out expired items, put older items in front.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[TURNTO] <food_bread>', '[LOOKAT] <food_bread>', '[PUTOBJBACK] <food_bread>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[FIND] <garbage_can>', '[PUTBACK] <food_cheese> <garbage_can>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[PUTOBJBACK] <food_cheese>', '[FIND] <food_dessert>', '[TURNTO] <food_dessert>', '[LOOKAT] <food_dessert>', '[PUSH] <food_dessert>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <milk>', '[GRAB] <milk>', '[TURNTO] <milk>', '[LOOKAT] <milk>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['chair', 'dough', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'tvstand', 'stereo', 'mousepad', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'sink', 'bathtub', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to the bedroom and sit at my desk. I turn my radio on and listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'bathroom', 'detergent', 'food_carrot', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'thread', 'stereo', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'cup', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'hair', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, turn to the sofa. Sit on the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'mop', 'microwave', 'trashcan', 'drawing', 'ceiling', 'character', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'tvstand', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'form', 'pillow', 'nightstand', 'newspaper', 'bed', 'curtain', 'towel_rack', 'dishrack', 'home_office', 'light', 'dishwasher', 'table', 'scrabble']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dishrack>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <dishrack> <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'food_ice_cream', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'toaster', 'stovefan', 'mat', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'wall', 'oven', 'powersocket', 'cutting_board', 'coffe_maker', 'television', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'hair', 'door', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I enter the kitchen and grab a waterglass from the cupboard. I open the fridge and grap some water. I pour the water into the cup and drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[OPEN] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'ceilinglamp', 'faucet', 'shower', 'towel', 'dining_room', 'wall', 'powersocket', 'bed', 'check', 'wooden_spoon', 'cards', 'mouse', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom. turn on the water in the sink. apply soap to hands. rub hands together. rinse hands with water. turn off the water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'cat', 'food_butter', 'hanger', 'floor', 'ceiling', 'food_cake', 'ceilinglamp', 'deck_of_cards', 'tvstand', 'homework', 'mousepad', 'closetdrawer', 'teeth', 'food_jam', 'nightstand', 'powersocket', 'bed', 'check', 'video_game_controller', 'mouthwash', 'towel_rack', 'mouse', 'console', 'cd', 'cupboard', 'instrument_guitar', 'coffee', 'shoe_rack']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Clean the welcome mat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <vacuum_cleaner>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'orchid', 'drawing', 'hanger', 'floor', 'phone', 'character', 'keyboard', 'toaster', 'plate', 'faucet', 'shower', 'window', 'mousepad', 'closetdrawer', 'dining_room', 'oven', 'powersocket', 'cup', 'bed', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'dresser', 'sink', 'bathtub', 'home_office', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'toilet', 'freezer', 'ground_coffee', 'bathroom', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'bookshelf', 'alcohol', 'mat', 'shower', 'window', 'dining_room', 'food_kiwi', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'television', 'doorjamb', 'mouse', 'door', 'light', 'colander', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I get my groceries and go to the fridge. I open the fridge and put the groceries there.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'desk', 'computer', 'ceilinglamp', 'shower', 'window', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'bed', 'food_food', 'mouse', 'vacuum_cleaner', 'bathtub', 'sink', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the living room, walk over to phone, grab the phone, put it to your ear .", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'mop', 'toilet', 'freezer', 'walllamp', 'bathroom', 'cat', 'food_butter', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'computer', 'mat', 'tvstand', 'iron', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'toy', 'tape', 'photoframe', 'television', 'food_rice', 'doorjamb', 'cards', 'door']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'stovefan', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'tape', 'newspaper', 'bed', 'television', 'doorjamb', 'mouse', 'sink', 'home_office', 'table', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office. Sit on chair. Switch on computer. type in keyboard.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'freezer', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'remote_control', 'computer', 'button', 'toaster', 'mat', 'shower', 'food_peanut_butter', 'food_kiwi', 'nightstand', 'oven', 'photoframe', 'television', 'oven_mitts', 'tray', 'kettle', 'cards', 'pencil', 'kitchen_counter', 'towel_rack', 'bathtub', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk into living room. Find television remote control. Pick up remote control. Press button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['food_bread', 'orchid', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bench', 'food_food', 'water', 'doorjamb', 'towel_rack', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I turn on faucet. I hold out my hands. I get my hands wet.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'food_bacon', 'alcohol', 'ceilinglamp', 'pillow', 'wall', 'powersocket', 'bed', 'curtain', 'doorjamb', 'kettle', 'mouse', 'dresser', 'cupboard', 'door', 'maindoor', 'light', 'creditcard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I open the room door and press the light button on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'stereo', 'couch', 'wall', 'nightstand', 'bench', 'powersocket', 'tape', 'bed', 'coffe_maker', 'check', 'tray', 'curtain', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[PUSH] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'chair', 'bathroom', 'detergent', 'trashcan', 'toothbrush', 'knifeblock', 'bathroom_counter', 'character', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'nightstand', 'oven', 'bench', 'powersocket', 'cup', 'bed', 'ice', 'diary', 'cards', 'dresser', 'hairbrush', 'home_office', 'dishwasher', 'scrabble']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": go to radio. run on radio.,\n    \"actions\": ['[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'walllamp', 'cat', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'toaster', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'oven', 'powersocket', 'bed', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'kitchen_counter', 'home_office', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "The cat is laying on the couch in the living room. I want to pet it, so I am going to walk to the living room, sit on the couch, and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'toilet', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'phone', 'character', 'piano_bench', 'bookmark', 'toaster', 'mat', 'shower', 'dish_soap', 'couch', 'wall', 'photoframe', 'novel', 'coffe_maker', 'check', 'cards', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'sink', 'home_office', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": walk to kitchen, find the dishwasher, apply dish soap into vessels or dishwasher, wash the dishes with a dishwasher,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[OPEN] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[WASH] <dishwasher>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'cat', 'drawing', 'cpuscreen', 'floor', 'bedroom', 'keyboard', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'check', 'coffe_maker', 'television', 'curtain', 'cupboard', 'food_steak', 'sink', 'home_office', 'table', 'light', 'instrument_guitar']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to living room.  Find sofa.  Sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'chair', 'jelly', 'walllamp', 'cat', 'trashcan', 'hanger', 'floor', 'ceiling', 'phone', 'food_bacon', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'couch', 'wall', 'oven', 'bed', 'tray', 'kitchen_counter', 'shoes', 'dishrack', 'dresser', 'cupboard', 'bathroom_cabinet', 'light', 'teeth']\nTask:Put away shoes\nresponse:\n{\n    \"thoughts\": Pick up shoes from entry hall. Open entry way closet. Put shoes in closet. Close closet door.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[FIND] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'walllamp', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'faucet', 'shower', 'towel', 'pillow', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'board_game', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom, open the faucet rinse the hands, wash the hands with soap and rinse again, then wipe with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[TURNTO] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'drawing', 'clothes_underwear', 'hanger', 'bathroom_counter', 'bedroom', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mat', 'mousepad', 'couch', 'pillow', 'dining_room', 'food_food', 'check', 'coffe_maker', 'mirror', 'ice', 'wooden_spoon', 'rag', 'cupboard', 'door', 'sink', 'table']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": Spray glass cleaner on mirror. Wipe with a clean cloth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[WALK] <mirror>', '[PUTOBJBACK] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'knifeblock', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'character', 'bookshelf', 'brush', 'ceilinglamp', 'mat', 'stereo', 'toaster', 'pillow', 'wall', 'bench', 'photoframe', 'bed', 'novel', 'wine', 'after_shave', 'vacuum_cleaner', 'dresser', 'pajamas', 'table', 'board_game', 'shoe_rack']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug in vacuum. Walk around with vacuum, covering all of the floor. Unplug vacuum.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'detergent', 'orchid', 'clothes_underwear', 'floor', 'character', 'button', 'stovefan', 'ceilinglamp', 'toaster', 'homework', 'mousepad', 'couch', 'dustpan', 'microphone', 'form', 'towel', 'powersocket', 'basket_for_clothes', 'bed', 'light_bulb', 'oven_mitts', 'tray', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'colander', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to laundry room. Find laundry basket. Grab laundry basket. Find washing machine. Walk to washing machine. Open washing machine. Find detergent. Grab detergent. Put in detergent. Put in clothes. Close washing machine. Turn on machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[PUTOBJBACK] <laundry_detergent>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'phone', 'remote_control', 'food_bacon', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'powersocket', 'oven', 'cup', 'bed', 'coffe_maker', 'television', 'curtain', 'lighter', 'doorjamb', 'shaving_cream', 'headset', 'cupboard', 'table', 'clothes_shirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "We recently got a 43 inch TV and have enjoyed watching some of our favorite movies and TV shows. We use a remote to do so.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'detergent', 'food_butter', 'clothes_underwear', 'hanger', 'floor', 'stamp', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'shoe-shine_kit', 'clothes_socks', 'closetdrawer', 'couch', 'clothes_pants', 'basket_for_clothes', 'check', 'mouthwash', 'food_rice', 'vacuum_cleaner', 'laundry_detergent', 'clothes_dress', 'bathroom_cabinet', 'light', 'dishwasher', 'clothes_shirt', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[DROP] <basket_for_clothes>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[GRAB] <laundry_detergent>', '[PUTOBJBACK] <laundry_detergent>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sponge', 'scissors', 'food_salt', 'trashcan', 'drawing', 'needle', 'hanger', 'longboard', 'floor', 'character', 'stamp', 'keyboard', 'stovefan', 'ceilinglamp', 'deck_of_cards', 'window', 'couch', 'dining_room', 'wall', 'photoframe', 'television', 'curtain', 'rag', 'pencil', 'shoes', 'clothes_dress', 'conditioner', 'sink', 'table']\nTask:Text friends while sitting on couch\nresponse:\n", "output": {"thoughts": "Text while sitting", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[TOUCH] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'mop', 'bathroom', 'cat', 'trashcan', 'knifeblock', 'stamp', 'button', 'piano_bench', 'faucet', 'mousepad', 'bowl', 'nightstand', 'wall_clock', 'oven', 'wall', 'bench', 'photoframe', 'novel', 'food_food', 'coffe_maker', 'curtain', 'pencil', 'shoes', 'hairbrush', 'hair', 'instrument_guitar', 'dishwasher', 'tooth_paste']\nTask:Read book\nresponse:\n{\n    \"thoughts\": pick up  the book open the book and read,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'foundation', 'couch', 'wall', 'pasta', 'toy', 'box', 'television', 'curtain', 'diary', 'towel_rack', 'kitchen_counter', 'cupboard', 'clothes_scarf', 'home_office', 'table', 'vase', 'coffee']\nTask:Change TV channels\nresponse:\n", "output": {"thoughts": "Take remote. Change to desired channel by you.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[TOUCH] <remote_control>", "[TOUCH] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'ground_coffee', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'crayon', 'microwave', 'toilet', 'detergent', 'milk', 'folder', 'bedroom', 'toaster', 'stovefan', 'balanceball', 'stereo', 'window', 'mousepad', 'food_noodles', 'dining_room', 'nightstand', 'box', 'newspaper', 'tray', 'lighter', 'headset', 'controller', 'door', 'maindoor', 'instrument_guitar', 'table', 'soap']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'blow_dryer', 'trashcan', 'drawing', 'knife', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'toaster', 'faucet', 'picture', 'towel', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'novel', 'food_food', 'controller', 'door', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'soap', 'board_game']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. Walk to the sink. Turn on tap. Get soap. Wash and rinse hands. Turn off tap. Dry hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[LOOKAT] <sink>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[RINSE] <soap>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'toilet', 'freezer', 'bathroom', 'trashcan', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'keyboard', 'remote_control', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'food_jam', 'pillow', 'wall', 'television', 'curtain', 'towel_rack', 'chessboard', 'table', 'light']\nTask:Watch  TV\nresponse:\n", "output": {"thoughts": "Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'detergent', 'orchid', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'purse', 'envelope', 'mat', 'window', 'foundation', 'broom', 'wall', 'nightstand', 'keys', 'novel', 'check', 'toothbrush_holder', 'oven_mitts', 'curtain', 'doorjamb', 'mouse', 'dresser', 'hairdryer', 'home_office', 'clothes_shirt', 'coffee', 'creditcard']\nTask:Put on coat\nresponse:\n{\n    \"thoughts\": Open closet and grab coat.  Put on coat and fasten the front closed.  Close closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'tablelamp', 'crayon', 'toilet', 'bathroom', 'cat', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'faucet', 'shower', 'oil', 'clothes_pants', 'pillow', 'wall', 'bench', 'photoframe', 'television', 'shaving_cream', 'shoes', 'sink', 'bathroom_cabinet', 'light']\nTask:Clean sink\nresponse:\n", "output": {"thoughts": "Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.", "actions": ["[WALK] <bathroom>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[WIPE] <faucet>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'ceilingfan', 'drawing', 'floor', 'bookshelf', 'ceilinglamp', 'shower', 'window', 'couch', 'food_orange', 'towel', 'dining_room', 'wall', 'powersocket', 'television', 'tray', 'wooden_spoon', 'rag', 'lighter', 'towel_rack', 'pencil', 'dresser', 'clothes_dress', 'hairdryer', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'clothes_shirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, find sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'measuring_cup', 'chair', 'ground_coffee', 'trashcan', 'orchid', 'drawing', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'powersocket', 'cup', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'maindoor', 'light', 'table']\nTask:Hang pictures\nresponse:\n", "output": {"thoughts": "My daughter was a week early which threw us for a loop. I'm still hanging pictures in her nursery.", "actions": ["[WALK] <bedroom>", "[WALK] <drawing>", "[FIND] <drawing>", "[GRAB] <drawing>", "[FIND] <wall>", "[PUTBACK] <drawing> <wall>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[FIND] <drawing>", "[GRAB] <drawing>", "[PUTBACK] <drawing> <wall>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'milk', 'toothbrush', 'drawing', 'orchid', 'cpuscreen', 'character', 'bedroom', 'stovefan', 'toaster', 'mousepad', 'clothes_pants', 'food_kiwi', 'wall_clock', 'bed', 'television', 'toothbrush_holder', 'water', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'shoes', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'dirt']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <toothbrush>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'microwave', 'walllamp', 'food_butter', 'knifeblock', 'drawing', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'tvstand', 'balanceball', 'picture', 'shower', 'oil', 'food_peanut_butter', 'pillow', 'wall', 'oven', 'television', 'coin', 'diary', 'mouse', 'cupboard', 'bag', 'bathroom_cabinet', 'vase', 'shoe_rack']\nTask:Make popcorn\nresponse:\n", "output": {"thoughts": "Lay popcorn bag flat inside microwave and press popcorn button to start.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <microwave>", "[TURNTO] <microwave>", "[OPEN] <microwave>", "[PUTIN] <food_food> <microwave>", "[CLOSE] <microwave>", "[SWITCHON] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'band-aids', 'toothbrush', 'orchid', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'faucet', 'picture', 'shower', 'mousepad', 'candle', 'food_orange', 'oven', 'basket_for_clothes', 'bed', 'check', 'food_food', 'television', 'tray', 'after_shave', 'headset', 'comb', 'dresser', 'laundry_detergent', 'napkin', 'bathroom_cabinet']\nTask:Sit in chair\nresponse:\n{\n    \"thoughts\": I sit in a chair to rest and eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'mail', 'toaster', 'mat', 'faucet', 'stereo', 'window', 'wall', 'oven', 'powersocket', 'cup', 'photoframe', 'television', 'tray', 'food_onion', 'centerpiece', 'hairbrush', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet', 'vase', 'soap', 'tooth_paste', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap and hot water", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'microwave', 'toilet', 'freezer', 'trashcan', 'food_carrot', 'drawing', 'floor', 'ceiling', 'mat', 'ceilinglamp', 'shower', 'couch', 'wall', 'novel', 'food_food', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'home_office', 'instrument_guitar', 'bathroom_cabinet', 'table', 'dirt']\nTask:Added meat to freezer\nresponse:\n{\n    \"thoughts\": After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'nail_polish', 'food_carrot', 'dough', 'knife', 'drawing', 'floor', 'ceiling', 'character', 'desk', 'alcohol', 'table_cloth', 'mat', 'tvstand', 'window', 'couch', 'food_orange', 'wall', 'oven', 'photoframe', 'check', 'television', 'doorjamb', 'console', 'food_steak', 'sink', 'bathroom_cabinet', 'table', 'home_office']\nTask:Put out flowers\nresponse:\n", "output": {"thoughts": "Place a flower centrepiece on the entrance hall table.", "actions": ["[WALK] <home_office>", "[WALK] <centerpiece>", "[FIND] <centerpiece>", "[GRAB] <centerpiece>", "[WALK] <table>", "[PUTBACK] <centerpiece> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'hanger', 'bathroom_counter', 'character', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'plate', 'tvstand', 'faucet', 'shower', 'dish_soap', 'mousepad', 'wall', 'wall_clock', 'oven', 'photoframe', 'bed', 'television', 'tray', 'towel_rack', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": walk into kitchen, find dishwasher, put plates in dishwasher, put soap in dishwasher, turn on dishwasher,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['chair', 'drawing', 'hanger', 'floor', 'phone', 'bedroom', 'purse', 'toaster', 'laptop', 'plate', 'ceilinglamp', 'tooth_paste', 'form', 'food_kiwi', 'wall', 'bench', 'powersocket', 'photoframe', 'food_food', 'television', 'centerpiece', 'doorjamb', 'diary', 'dresser', 'cupboard', 'door', 'maindoor', 'table', 'light', 'board_game']\nTask:Eat\nresponse:\n", "output": {"thoughts": "Eating a meal", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <plate>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <plate>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'juice', 'cd_player', 'toilet', 'freezer', 'trashcan', 'bedroom', 'remote_control', 'button', 'clothes_jacket', 'toaster', 'laptop', 'ceilinglamp', 'plate', 'homework', 'dining_room', 'broom', 'bench', 'tape', 'keys', 'coffe_maker', 'television', 'towel_rack', 'hair', 'cupboard', 'maindoor', 'sink', 'clothes_skirt', 'vase']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cd_player>', '[FIND] <cd_player>', '[PLUGIN] <cd_player>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['fork', 'drawing', 'cpuscreen', 'ceiling', 'character', 'bedroom', 'man', 'keyboard', 'table_cloth', 'brush', 'stovefan', 'balanceball', 'picture', 'mousepad', 'couch', 'bowl', 'nightstand', 'pasta', 'cup', 'food_cheese', 'photoframe', 'curtain', 'shaving_cream', 'cards', 'laundry_detergent', 'maindoor', 'sink', 'napkin', 'home_office', 'bathroom_cabinet']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Wait until doorbell rings, look to see who it is outside and open the door if you feel so inclined.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'dry_pasta', 'tablelamp', 'walllamp', 'detergent', 'milk', 'orchid', 'floor', 'ceiling', 'button', 'toaster', 'faucet', 'wall', 'box', 'basket_for_clothes', 'keys', 'coffe_maker', 'food_food', 'curtain', 'food_sugar', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'bathtub', 'sink', 'home_office', 'coffee_cup', 'dishwasher', 'clothes_skirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[LOOKAT] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <milk>', '[GRAB] <milk>', '[PUTBACK] <milk> <coffe_maker>', '[FIND] <food_sugar>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[SWITCHON] <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'phone', 'mop_bucket', 'bookmark', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'mat', 'stovefan', 'window', 'iron', 'couch', 'wall', 'food_cheese', 'photoframe', 'coffee_table', 'novel', 'video_game_controller', 'television', 'mouse', 'maindoor', 'bag', 'home_office', 'table', 'clothes_shirt']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit down in recliner.  Pick up novel off of coffee table.  Open novel to last read page.  Begin reading", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bills', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'television', 'toothbrush_holder', 'wine', 'doorjamb', 'mouse', 'dresser', 'hairbrush', 'controller', 'clothes_dress', 'home_office', 'light', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'ceilingfan', 'food_carrot', 'toothbrush', 'drawing', 'knife', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'cup', 'photoframe', 'light_bulb', 'coffe_maker', 'television', 'tray', 'curtain', 'pencil', 'maindoor', 'table', 'light', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn TV on to TV channel and sit down on the couch and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I want to surf the internet and look around at funny pictures.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'dog', 'freezer', 'dough', 'knifeblock', 'knife', 'hanger', 'bathroom_counter', 'ceiling', 'toaster', 'tvstand', 'faucet', 'couch', 'pasta', 'powersocket', 'oven', 'photoframe', 'novel', 'television', 'wooden_spoon', 'curtain', 'coffee_pot', 'mouse', 'dresser', 'door', 'sink', 'home_office', 'light', 'bathroom_cabinet']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Say hi", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'chair', 'freezer', 'nail_polish', 'bathroom', 'food_butter', 'knifeblock', 'cpuscreen', 'hanger', 'ceiling', 'dvd_player', 'stamp', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'iron', 'nightstand', 'oven', 'cutting_board', 'check', 'television', 'coin', 'tray', 'doorjamb', 'bathroom_cabinet', 'teeth']\nTask:Answer door\nresponse:\n{\n    \"thoughts\": Walk up to the door.  Open the door and greet whomever is out side,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['pot', 'mop', 'toilet', 'chair', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'computer', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'wall_clock', 'bench', 'television', 'wine', 'curtain', 'coffee_pot', 'door', 'sink', 'spoon', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I watch Netflix from about 8:30 to 10:30 every night. I usually watch TV shows.", "actions": ["[WALK] <home_office>", "[WALK] <wall_clock>", "[FIND] <wall_clock>", "[TURNTO] <wall_clock>", "[LOOKAT] <wall_clock>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'band-aids', 'trashcan', 'dough', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'dvd_player', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'mousepad', 'food_orange', 'wall', 'oven', 'keys', 'mouse', 'shoes', 'hair', 'dishwasher', 'vase', 'coffee']\nTask:Read book\nresponse:\n{\n    \"thoughts\": find a book to read. sit down. read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'bathroom', 'ceilingfan', 'milk', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'shoe-shine_kit', 'couch', 'food_jam', 'microphone', 'wall', 'nightstand', 'oven', 'cutting_board', 'television', 'curtain', 'pencil', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "We sit all around the television and watch a program.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'mop', 'freezer', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'table_cloth', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'keys', 'newspaper', 'food_rice', 'doorjamb', 'dresser', 'door', 'bathroom_cabinet', 'light', 'scrabble']\nTask:Turn light off\nresponse:\n{\n    \"thoughts\": Put the light switch into the down, or off, position.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'toilet', 'chair', 'walllamp', 'trashcan', 'food_carrot', 'clothes_underwear', 'hanger', 'longboard', 'floor', 'ceiling', 'toaster', 'electric_shaver', 'couch', 'wall', 'nightstand', 'oven', 'cup', 'coffe_maker', 'food_food', 'television', 'cards', 'mouse', 'kitchen_counter', 'maindoor', 'home_office', 'light', 'dishwasher', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'food_ice_cream', 'toothbrush', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'purse', 'faucet', 'window', 'towel', 'wall', 'toy', 'bed', 'light_bulb', 'novel', 'coffe_maker', 'television', 'water', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'dresser', 'maindoor', 'sink', 'tooth_paste', 'clothes_shirt', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[FIND] <toothbrush>', '[FIND] <water>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['spectacles', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'character', 'bookshelf', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'shower', 'stereo', 'couch', 'wall', 'cutting_board', 'cup', 'food_cheese', 'novel', 'food_food', 'television', 'wine', 'curtain', 'door', 'table']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sit down on couch in front of TV.  Use remote to turn on the TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'toilet', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'envelope', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'stereo', 'mousepad', 'towel', 'wall', 'powersocket', 'photoframe', 'bed', 'toilet_paper', 'tray', 'curtain', 'doorjamb', 'dresser', 'controller', 'door', 'food_steak', 'light', 'tooth_paste']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": walk to bathroom , find shower , take a shower,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'crayon', 'tea', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'mat', 'faucet', 'picture', 'wall', 'oven', 'food_food', 'check', 'television', 'ice', 'tray', 'rag', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'spoon']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Take out a rag. Turn on faucet and get water on the rag. Clean counter off.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <rag> <sink>", "[SWITCHOFF] <faucet>", "[GRAB] <rag>", "[SQUEEZE] <rag>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'bathroom', 'food_butter', 'milk', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'bookshelf', 'computer', 'ceilinglamp', 'shower', 'window', 'mousepad', 'closetdrawer', 'couch', 'clothes_hat', 'wall', 'coffe_maker', 'television', 'tray', 'curtain', 'dresser', 'bathtub', 'bathroom_cabinet', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to the entrance hall. Find the telephone. Grab the telephone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['juice', 'measuring_cup', 'microwave', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'face', 'mop_bucket', 'tvstand', 'faucet', 'food_peanut_butter', 'towel', 'wall', 'oven', 'face_soap', 'tape', 'photoframe', 'bed', 'keys', 'rag', 'cards', 'comb', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <rag>", "[GRAB] <rag>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[WASH] <rag>", "[FIND] <face_soap>", "[GRAB] <face_soap>", "[POUR] <face_soap> <rag>", "[PUTOBJBACK] <face_soap>", "[FIND] <face>", "[SCRUB] <face>", "[PUTBACK] <rag> <sink>", "[RINSE] <face>", "[RINSE] <rag>", "[SWITCHOFF] <faucet>", "[GRAB] <rag>", "[PUTOBJBACK] <rag>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <face>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'blow_dryer', 'bathroom', 'ceilingfan', 'drawing', 'hanger', 'kitchen_cabinet', 'floor', 'phone', 'desk', 'computer', 'laptop', 'plate', 'stereo', 'oil', 'clothes_hat', 'wall', 'bench', 'oven', 'powersocket', 'newspaper', 'coffe_maker', 'curtain', 'rag', 'kitchen_counter', 'maindoor', 'home_office', 'table', 'light', 'soap']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk into kitchen. Grab plates and forks. Walk into dining room. Put plates and forks on table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[PUTBACK] <fork> <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'beer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'shower', 'couch', 'pillow', 'wall', 'powersocket', 'tape', 'photoframe', 'after_shave', 'doorjamb', 'shoes', 'comb', 'door', 'maindoor', 'bathroom_cabinet', 'light', 'vase', 'colander', 'coffee']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'cat', 'dough', 'bathroom_counter', 'phone', 'bookshelf', 'ceilinglamp', 'tvstand', 'stereo', 'candle', 'bowl', 'dining_room', 'wall', 'newspaper', 'coffe_maker', 'food_food', 'coffee_filter', 'tray', 'curtain', 'diary', 'towel_rack', 'pencil', 'dresser', 'food_steak', 'sink', 'napkin', 'door', 'dishwasher']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[WALK] <phone>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'faucet', 'picture', 'window', 'dustpan', 'towel', 'wall', 'nightstand', 'photoframe', 'keys', 'toothbrush_holder', 'food_donut', 'mouse', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Cleaning hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_butter', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'toaster', 'stereo', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'bench', 'cutting_board', 'cup', 'check', 'food_donut', 'food_rice', 'curtain', 'centerpiece', 'mouse', 'vacuum_cleaner', 'shoes', 'dresser', 'light', 'teeth']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Run the vacuum cleaner across the floor, back and forth a couple of times until clean.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'microwave', 'cat', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'character', 'desk', 'mat', 'deck_of_cards', 'iron', 'stereo', 'balanceball', 'shower', 'faucet', 'mousepad', 'nightstand', 'oven', 'television', 'tray', 'curtain', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'clothes_dress', 'sink', 'vase']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'crayon', 'ceilingfan', 'drawing', 'needle', 'floor', 'ceiling', 'shampoo', 'keyboard', 'remote_control', 'thread', 'food_bacon', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'photoframe', 'bed', 'television', 'coffee_filter', 'curtain', 'mouse', 'table', 'light', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go into the living room. I find the remote. I turn on the tv. I sit on the couch and watch tv.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'walllamp', 'knife', 'knifeblock', 'cpuscreen', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'tea_bag', 'toaster', 'stovefan', 'plate', 'faucet', 'deck_of_cards', 'tvstand', 'bowl', 'wall', 'oven', 'cup', 'coffe_maker', 'drinking_glass', 'food_donut', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Unload dishwasher\nresponse:\n", "output": {"thoughts": "Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[PUTBACK] <drinking_glass> <cupboard>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[PUTIN] <fork> <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTIN] <knife> <kitchen_cabinet>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTIN] <spoon> <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTIN] <plate> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'bathroom', 'hanger', 'floor', 'phone', 'ceiling', 'character', 'bedroom', 'bookshelf', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'closetdrawer', 'couch', 'pillow', 'cup', 'coffe_maker', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'home_office', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_bread', 'freezer', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'button', 'stovefan', 'toaster', 'ceilinglamp', 'faucet', 'wall', 'oven', 'food_food', 'video_game_controller', 'television', 'razor', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'sink', 'spoon', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Cooking a meal", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <pot> <sink>", "[GRAB] <pot>", "[SWITCHOFF] <faucet>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[CLOSE] <freezer>", "[WALK] <oven>", "[PUTBACK] <food_food> <pot>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'microwave', 'walllamp', 'detergent', 'drawing', 'wallshelf', 'floor', 'ceiling', 'computer', 'button', 'food_bacon', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'coffee_table', 'novel', 'mouthwash', 'television', 'doorjamb', 'diary', 'clothes_scarf', 'sink']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit down in recliner.  Pick up novel off of coffee table.  Open novel to last read page.  Begin reading,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <bookmark>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'scissors', 'juice', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'faucet', 'towel', 'dining_room', 'wall', 'photoframe', 'food_food', 'razor', 'television', 'food_rice', 'wooden_spoon', 'after_shave', 'headset', 'laundry_detergent', 'conditioner', 'sink', 'bathroom_cabinet', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Turn on the water faucet. Wet my hands and turn off the faucet. Put soap on my hands. Rub hands together, making sure the soap covers all areas of the hands. Scrub hands for about 20 seconds. Turn on the water water faucet. Rinse hands under the running water. Turn off the water faucet. Dry hands with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'ceilinglamp', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'mouse', 'chessboard', 'dresser', 'cupboard', 'door', 'light', 'table']\nTask:Re arrange office\nresponse:\n{\n    \"thoughts\": Walk in.Turn on light.Move furniture. Change the room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[PUSH] <desk>', '[WALK] <chair>', '[PUSH] <chair>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['scissors', 'bathroom', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bookshelf', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'couch', 'dining_room', 'food_kiwi', 'wall', 'toilet_paper', 'razor', 'television', 'curtain', 'diary', 'towel_rack', 'door', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch and use the remote control to turn on the TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'measuring_cup', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'doorjamb', 'mouse', 'pencil', 'cupboard', 'door', 'maindoor', 'home_office', 'light', 'dishwasher', 'creditcard']\nTask:Work\nresponse:\n{\n    \"thoughts\": turn computer on, open application to work with,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'microwave', 'ceilingfan', 'orchid', 'bathroom_counter', 'bookshelf', 'purse', 'stovefan', 'mat', 'iron', 'electric_shaver', 'food_noodles', 'mousepad', 'couch', 'clothes_pants', 'pillow', 'powersocket', 'keys', 'coffe_maker', 'oven_mitts', 'wine', 'shaving_cream', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'home_office', 'light', 'tooth_paste']\nTask:Hang up car keys\nresponse:\n", "output": {"thoughts": "Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <purse>", "[FIND] <purse>", "[OPEN] <purse>", "[FIND] <keys>", "[GRAB] <keys>", "[CLOSE] <purse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'tea', 'ground_coffee', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'button', 'mat', 'deck_of_cards', 'plate', 'tvstand', 'closetdrawer', 'couch', 'food_jam', 'dining_room', 'wall', 'powersocket', 'basket_for_clothes', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'chessboard', 'door', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk into living room, walk to light, turn on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'juice', 'scissors', 'beer', 'chair', 'freezer', 'bathroom', 'ceilingfan', 'hanger', 'floor', 'computer', 'food_bacon', 'mat', 'couch', 'creditcard', 'dining_room', 'oven', 'tape', 'bed', 'light_bulb', 'novel', 'food_food', 'mouse', 'cupboard', 'conditioner', 'home_office', 'table', 'scrabble', 'shoe_rack']\nTask:Pet dog\nresponse:\n", "output": {"thoughts": "Pet my dog in hall", "actions": ["[WALK] <home_office>", "[WALK] <dog>", "[TURNTO] <dog>", "[LOOKAT] <dog>", "[TOUCH] <dog>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'dog', 'blow_dryer', 'trashcan', 'orchid', 'bathroom_counter', 'longboard', 'ceiling', 'character', 'bedroom', 'computer', 'mop_bucket', 'stereo', 'shower', 'window', 'food_peanut_butter', 'food_orange', 'bench', 'oven', 'bed', 'toilet_paper', 'television', 'razor', 'curtain', 'towel_rack', 'cupboard', 'door', 'dishwasher', 'soap']\nTask:Put umbrella away\nresponse:\n{\n    \"thoughts\": Place the umbrella in the umbrella bucket point first.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mop_bucket>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'blender', 'food_salt', 'beer', 'tea', 'ceilingfan', 'toothbrush', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'face', 'mat', 'faucet', 'food_noodles', 'towel', 'wall', 'facial_cleanser', 'photoframe', 'food_rice', 'after_shave', 'rag', 'kitchen_counter', 'sink', 'napkin', 'bathroom_cabinet', 'dirt']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Turn the faucet on. Wet the face. Take facial cleanser out of packaging. Lather on face. Rinse face with water. Turn off the faucet. Dry face with towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <face>", "[WASH] <face>", "[FIND] <rag>", "[GRAB] <rag>", "[RINSE] <rag>", "[FIND] <facial_cleanser>", "[GRAB] <facial_cleanser>", "[POUR] <facial_cleanser> <rag>", "[PUTOBJBACK] <facial_cleanser>", "[SCRUB] <face>", "[RINSE] <rag>", "[PUTOBJBACK] <rag>", "[RINSE] <face>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <face>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['drawing', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'computer', 'button', 'mat', 'ceilinglamp', 'shower', 'faucet', 'tvstand', 'food_peanut_butter', 'towel', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'curtain', 'mouse', 'kitchen_counter', 'shoes', 'laundry_detergent', 'food_steak', 'sink', 'bag', 'bathroom_cabinet', 'door', 'vase', 'soap']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": I want to take a shower so I go to the bathroom, turn on the shower, get the soap, lather and wash, then I rinse, turn off shower, get a towel and dry off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[RINSE] <shower>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'toilet', 'chair', 'walllamp', 'cpuscreen', 'longboard', 'floor', 'character', 'desk', 'bookshelf', 'keyboard', 'envelope', 'computer', 'button', 'table_cloth', 'ceilinglamp', 'window', 'mousepad', 'oil', 'dustpan', 'wall', 'nightstand', 'powersocket', 'photoframe', 'clothes_gloves', 'mouse', 'sink', 'bathroom_cabinet', 'printing_paper']\nTask:Add paper to printer\nresponse:\n", "output": {"thoughts": "Take a stack of paper, about 25 pieces. Place carefully in the paper tray.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <printing_paper>", "[GRAB] <printing_paper>", "[WALK] <fax_machine>", "[PUTBACK] <printing_paper> <fax_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['nail_polish', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'faucet', 'couch', 'dining_room', 'wall', 'nightstand', 'bench', 'wall_clock', 'photoframe', 'coffe_maker', 'food_food', 'lighter', 'doorjamb', 'towel_rack', 'diary', 'shoes', 'comb', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom. turn on the water in the sink. apply soap to hands. rub hands together. rinse hands with water. turn off the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sheets', 'tablelamp', 'nail_polish', 'trashcan', 'ceilingfan', 'drawing', 'longboard', 'floor', 'ceiling', 'phone', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'photoframe', 'television', 'curtain', 'laundry_detergent', 'bathroom_cabinet', 'table', 'light', 'shoe_rack']\nTask:Straighten pictures on wall\nresponse:\n", "output": {"thoughts": "Move the pictures/paintings until they are straight.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PUSH] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'computer', 'mat', 'shoe-shine_kit', 'oil', 'pillow', 'wall', 'pasta', 'powersocket', 'cup', 'food_food', 'video_game_controller', 'toothbrush_holder', 'television', 'doorjamb', 'kettle', 'door', 'light', 'clothes_skirt', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on light in an area,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['fork', 'food_pizza', 'filing_cabinet', 'tablelamp', 'freezer', 'nail_polish', 'cat', 'food_carrot', 'drawing', 'orchid', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'stamp', 'ceilinglamp', 'plate', 'candle', 'dining_room', 'wall', 'bench', 'food_vegetable', 'newspaper', 'food_food', 'television', 'after_shave', 'cards', 'cd', 'clothes_dress', 'table']\nTask:Fix snack\nresponse:\n", "output": {"thoughts": "Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <kitchen_cabinet>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTBACK] <food_carrot> <plate>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[PUTBACK] <food_vegetable> <plate>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'measuring_cup', 'microwave', 'freezer', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'face', 'mop_bucket', 'tvstand', 'faucet', 'food_peanut_butter', 'towel', 'wall', 'oven', 'face_soap', 'tape', 'photoframe', 'bed', 'keys', 'rag', 'cards', 'comb', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet']\nTask:Wash face\nresponse:\n{\n    \"thoughts\": Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <rag>', '[GRAB] <rag>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[WASH] <rag>', '[FIND] <face_soap>', '[GRAB] <face_soap>', '[POUR] <face_soap> <rag>', '[PUTOBJBACK] <face_soap>', '[FIND] <face>', '[SCRUB] <face>', '[PUTBACK] <rag> <sink>', '[RINSE] <face>', '[RINSE] <rag>', '[SWITCHOFF] <faucet>', '[GRAB] <rag>', '[PUTOBJBACK] <rag>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <face>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['spectacles', 'toilet', 'walllamp', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'remote_control', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'wall', 'powersocket', 'food_food', 'television', 'curtain', 'towel_rack', 'headset', 'door', 'bathroom_cabinet', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "We sit all around the television and watch a program.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[PUTBACK] <remote_control> <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'measuring_cup', 'microwave', 'blow_dryer', 'bathroom', 'woman', 'knifeblock', 'clothes_underwear', 'phone', 'dvd_player', 'man', 'mail', 'toaster', 'iron', 'faucet', 'food_orange', 'child', 'powersocket', 'cup', 'food_cheese', 'check', 'tray', 'curtain', 'door', 'sink', 'home_office', 'dishwasher', 'clothes_skirt', 'tooth_paste']\nTask:Say goodbye to guests leaving\nresponse:\n{\n    \"thoughts\": Tell them bye,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[FIND] <man>', '[TOUCH] <man>', '[FIND] <woman>', '[TOUCH] <woman>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'trashcan', 'hanger', 'floor', 'ceiling', 'phone', 'mail', 'tvstand', 'clothes_socks', 'oil', 'mousepad', 'food_peanut_butter', 'food_orange', 'form', 'wall', 'cup', 'check', 'mouthwash', 'television', 'tray', 'doorjamb', 'dishrack', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'sink', 'light', 'soap']\nTask:Turn light off\nresponse:\n", "output": {"thoughts": "Put the light switch into the down, or off, position.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'walllamp', 'knifeblock', 'orchid', 'floor', 'phone', 'ceiling', 'bedroom', 'keyboard', 'food_bacon', 'stovefan', 'toaster', 'plate', 'faucet', 'window', 'homework', 'closetdrawer', 'wall', 'oven', 'ice', 'tray', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'hairdryer', 'sink', 'bathroom_cabinet', 'coffee']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[TURNTO] <plate>', '[LOOKAT] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WASH] <plate>', '[PUTBACK] <plate> <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'fork', 'tablelamp', 'food_salt', 'microwave', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'nightstand', 'bed', 'novel', 'television', 'curtain', 'lighter', 'mouse', 'door', 'table', 'dishwasher']\nTask:Admire art\nresponse:\n", "output": {"thoughts": "Looking at painting, art", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'mop', 'tea', 'freezer', 'walllamp', 'floor', 'character', 'stovefan', 'toaster', 'mat', 'tvstand', 'stereo', 'window', 'electric_shaver', 'food_noodles', 'couch', 'soap', 'pillow', 'wall', 'bench', 'toilet_paper', 'food_food', 'doorjamb', 'diary', 'dresser', 'hairbrush', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to bathroom. find closet. Uset It. Clean it. Use Toilet paper,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[CLOSE] <dresser>', '[FIND] <toilet_paper>', '[WASH] <dresser>', '[FIND] <toilet_paper>', '[TOUCH] <toilet_paper>']\n}\n", "input": "\nScene_Objects:['fork', 'spectacles', 'tea', 'ceilingfan', 'trashcan', 'drawing', 'food_egg', 'floor', 'ceiling', 'desk', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'oil', 'couch', 'pillow', 'wall', 'bed', 'coffe_maker', 'television', 'curtain', 'sink', 'bathroom_cabinet', 'table', 'light', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[PUTOBJBACK] <remote_control>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'detergent', 'cpuscreen', 'ceiling', 'computer', 'toaster', 'brush', 'laptop', 'faucet', 'mousepad', 'clothes_pants', 'dirt', 'powersocket', 'oven', 'cup', 'bed', 'food_food', 'television', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'chessboard', 'laundry_detergent', 'sink', 'dishwasher', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[PLUGIN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'woman', 'clothes_underwear', 'hanger', 'desk', 'man', 'bedroom', 'computer', 'ceilinglamp', 'mat', 'deck_of_cards', 'child', 'bench', 'cup', 'newspaper', 'photoframe', 'coffe_maker', 'video_game_controller', 'towel_rack', 'pencil', 'kitchen_counter', 'dresser', 'food_steak', 'sink', 'bathroom_cabinet', 'home_office', 'table', 'light', 'tooth_paste', 'laser_pointer']\nTask:Say goodbye to guests leaving\nresponse:\n", "output": {"thoughts": "Tell them bye", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[FIND] <man>", "[TOUCH] <man>", "[FIND] <woman>", "[TOUCH] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'measuring_cup', 'walllamp', 'bathroom', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'bookmark', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'bench', 'cup', 'photoframe', 'novel', 'check', 'coffe_maker', 'television', 'tray', 'curtain', 'doorjamb', 'bathroom_cabinet', 'light', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the book case, I select a book, I go back and sit down, then I read.,\n    \"actions\": ['[WALK] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'drawing', 'hanger', 'floor', 'phone', 'food_bacon', 'stovefan', 'plate', 'tvstand', 'faucet', 'couch', 'food_jam', 'wall', 'bench', 'powersocket', 'photoframe', 'food_food', 'oven_mitts', 'doorjamb', 'mouse', 'kitchen_counter', 'headset', 'cd', 'dresser', 'cupboard', 'clothes_dress', 'door', 'table', 'light']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "Sit down at the dinning table with my family and eat dinner.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[FIND] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[GRAB] <food_food>", "[FIND] <plate>", "[PUTBACK] <food_food> <plate>", "[GRAB] <food_food>", "[PUTOBJBACK] <food_food>", "[FIND] <fork>", "[GRAB] <fork>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'food_carrot', 'drawing', 'hanger', 'phone', 'ceiling', 'desk', 'computer', 'toaster', 'mat', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'towel', 'cup', 'food_cheese', 'keys', 'tray', 'after_shave', 'towel_rack', 'mouse', 'cupboard', 'sink', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'chair', 'drawing', 'bills', 'knife', 'floor', 'ceilinglamp', 'stereo', 'window', 'couch', 'food_jam', 'form', 'towel', 'dining_room', 'wall', 'basket_for_clothes', 'photoframe', 'novel', 'food_food', 'check', 'television', 'oven_mitts', 'razor', 'tray', 'pencil', 'dresser', 'pajamas', 'table', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit and read", "actions": ["[WALK] <home_office>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[CLOSE] <filing_cabinet>", "[WALK] <couch>", "[SIT] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'freezer', 'bathroom', 'trashcan', 'dough', 'knifeblock', 'orchid', 'floor', 'ceiling', 'clothes_jacket', 'brush', 'ceilinglamp', 'mat', 'balanceball', 'clothes_socks', 'couch', 'microphone', 'form', 'wall', 'food_food', 'check', 'oven_mitts', 'wine', 'mouse', 'cupboard', 'food_steak', 'spoon', 'table', 'tooth_paste']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Go kitchen and make something for eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>']\n}\n", "input": "\nScene_Objects:['sheets', 'crayon', 'food_butter', 'food_carrot', 'bills', 'hanger', 'longboard', 'phone', 'keyboard', 'computer', 'electric_shaver', 'window', 'homework', 'mousepad', 'wall', 'bench', 'powersocket', 'photoframe', 'newspaper', 'novel', 'television', 'tray', 'mouse', 'vacuum_cleaner', 'dresser', 'hairbrush', 'napkin', 'table', 'vase', 'electrical_outlet']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug vacuum into the wall. Turn on the vacuum. Move vacuum around the floor to clean it. Unplug vacuum.", "actions": ["[WALK] <home_office>", "[WALK] <vacuum_cleaner>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[FIND] <electrical_outlet>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[PUTOBJBACK] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'picture', 'mousepad', 'wall', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'cards', 'mouse', 'door', 'light', 'printing_paper']\nTask:Add paper to printer\nresponse:\n{\n    \"thoughts\": Take a stack of paper, about 25 pieces. Place carefully in the paper tray.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <printing_paper>', '[GRAB] <printing_paper>', '[WALK] <fax_machine>', '[PUTBACK] <printing_paper> <fax_machine>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'oil', 'couch', 'wall', 'cup', 'food_food', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'console', 'hairbrush', 'cupboard', 'food_steak', 'maindoor', 'door', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch favorite shows, netflix, hulu", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTBACK] <remote_control> <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'chair', 'walllamp', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'mousepad', 'pillow', 'wall', 'cutting_board', 'box', 'toilet_paper', 'doorjamb', 'centerpiece', 'towel_rack', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'door', 'light', 'dishwasher', 'shoe_rack']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'toilet', 'ceilingfan', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'character', 'video_game_console', 'food_cake', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'couch', 'wall', 'powersocket', 'tape', 'cup', 'video_game_controller', 'television', 'mouthwash', 'curtain', 'hairdryer', 'table', 'coffee']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'filing_cabinet', 'nail_polish', 'toothbrush', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'closetdrawer', 'pillow', 'wall', 'wall_clock', 'food_food', 'toothbrush_holder', 'tray', 'dresser', 'cupboard', 'controller', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'light', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I go into the bathroom, go to the sink, get toothbrush and toothpaste. Put toothpaste on the toothbrush, brush teeth, turn on faucet, rinse brush, and return toothbrush.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['food_salt', 'walllamp', 'detergent', 'orchid', 'drawing', 'cpuscreen', 'keyboard', 'envelope', 'computer', 'stovefan', 'faucet', 'electric_shaver', 'couch', 'candle', 'microphone', 'broom', 'nightstand', 'oven', 'bench', 'tape', 'cup', 'light_bulb', 'food_rice', 'wine', 'wooden_spoon', 'hair', 'maindoor', 'table', 'dishwasher', 'vase']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Get a broom or duster with long handle. Dust blades of ceiling fan.", "actions": ["[WALK] <bedroom>", "[WALK] <duster>", "[FIND] <duster>", "[GRAB] <duster>", "[PUTOBJBACK] <duster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'dresser', 'hairbrush', 'clothes_dress', 'door', 'sink', 'spoon', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['freezer', 'milk', 'drawing', 'orchid', 'knife', 'food_egg', 'hanger', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'folder', 'mail', 'ceilinglamp', 'tvstand', 'food_orange', 'wall', 'bench', 'cutting_board', 'food_vegetable', 'bed', 'food_food', 'television', 'tray', 'curtain', 'mouse', 'pencil', 'table', 'colander']\nTask:Chop vegetables\nresponse:\n", "output": {"thoughts": "Cut up veggies with knife", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <cutting_board>", "[GRAB] <cutting_board>", "[CLOSE] <kitchen_cabinet>", "[WALK] <table>", "[PUTBACK] <knife> <table>", "[PUTBACK] <cutting_board> <table>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <food_vegetable> <cutting_board>", "[GRAB] <knife>", "[CUT] <food_vegetable>", "[PUTOBJBACK] <knife>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'toaster', 'mat', 'tvstand', 'mousepad', 'pillow', 'dining_room', 'wall', 'coffee_table', 'novel', 'television', 'tray', 'curtain', 'doorjamb', 'diary', 'mouse', 'cupboard', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I walk to the living room and pick up a book from the coffee table. I find my favorite chair.  I sit in the armchair and open the book.  I read.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[WALK] <coffee_table>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_salt', 'ceilingfan', 'dough', 'drawing', 'orchid', 'longboard', 'floor', 'ceiling', 'phone', 'character', 'remote_control', 'mail', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'faucet', 'couch', 'wall', 'oven', 'keys', 'television', 'coin', 'coffee_filter', 'curtain', 'dresser', 'maindoor', 'home_office', 'table']\nTask:Watch  TV\nresponse:\n", "output": {"thoughts": "Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'trashcan', 'food_carrot', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'alcohol', 'button', 'ceilinglamp', 'plate', 'clothes_socks', 'couch', 'food_orange', 'wall', 'bench', 'powersocket', 'photoframe', 'light_bulb', 'food_food', 'coffee_filter', 'drinking_glass', 'water', 'towel_rack', 'dishrack', 'cupboard', 'clothes_dress', 'light', 'colander']\nTask:Drink\nresponse:\n{\n    \"thoughts\": im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'trashcan', 'hanger', 'floor', 'bedroom', 'computer', 'button', 'brush', 'mat', 'shower', 'faucet', 'window', 'clothes_hat', 'wall', 'child', 'oven', 'bed', 'newspaper', 'novel', 'curtain', 'centerpiece', 'doorjamb', 'mouse', 'cd', 'dresser', 'hairdryer', 'clothes_scarf', 'vase', 'board_game']\nTask:Story reading time\nresponse:\n", "output": {"thoughts": "Sit on bed with child, open mother goose, pick story, read book, point at the pretty pictures", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[SIT] <bed>", "[TURNTO] <child>", "[LOOKAT] <child>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'longboard', 'character', 'bedroom', 'computer', 'faucet', 'homework', 'clothes_socks', 'food_noodles', 'dining_room', 'photoframe', 'check', 'television', 'after_shave', 'rag', 'lighter', 'towel_rack', 'kitchen_counter', 'dresser', 'clothes_dress', 'home_office', 'light', 'laser_pointer', 'scrabble']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office. I find the telephone. I pick up the telephone handset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'toilet', 'bathroom', 'detergent', 'ceilingfan', 'food_butter', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'pillow', 'wall', 'food_food', 'television', 'tray', 'curtain', 'after_shave', 'cupboard', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "TURN ON TV, SIT ON COUCH ACROSS THE ROOM, USE REMOTE TO LOOK FOR SHOW TO WATCH", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'filing_cabinet', 'tablelamp', 'mop', 'toilet', 'walllamp', 'cat', 'drawing', 'floor', 'ceiling', 'character', 'bookshelf', 'food_cake', 'piano_bench', 'shower', 'faucet', 'pillow', 'wall', 'photoframe', 'bed', 'curtain', 'doorjamb', 'diary', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'soap', 'instrument_violin']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['chair', 'cat', 'trashcan', 'woman', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'mail', 'stovefan', 'oil', 'mousepad', 'clothes_hat', 'wall', 'child', 'powersocket', 'cutting_board', 'basket_for_clothes', 'television', 'curtain', 'doorjamb', 'chessboard', 'dresser', 'cupboard', 'pajamas', 'door', 'napkin', 'light']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Go to the kitchen, greet your family members", "actions": ["[WALK] <dining_room>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'hanger', 'floor', 'phone', 'desk', 'bookshelf', 'computer', 'piano_bench', 'stovefan', 'plate', 'tvstand', 'window', 'mousepad', 'wall', 'powersocket', 'box', 'photoframe', 'novel', 'doorjamb', 'towel_rack', 'dresser', 'cupboard', 'door', 'light', 'vase', 'colander', 'shoe_rack']\nTask:Read\nresponse:\n{\n    \"thoughts\": We pick out a book from the bookshelf and read it.,\n    \"actions\": ['[WALK] <novel>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'detergent', 'food_butter', 'dough', 'food_carrot', 'drawing', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'keyboard', 'computer', 'clothes_jacket', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'mousepad', 'towel', 'wall', 'oven', 'cup', 'photoframe', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap and hot water", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'measuring_cup', 'walllamp', 'bathroom', 'trashcan', 'ceilingfan', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'character', 'toaster', 'mat', 'dish_soap', 'couch', 'wall', 'nightstand', 'oven', 'powersocket', 'bench', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'shaving_cream', 'cupboard', 'food_steak', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I go to the dishwasher and put dish soap. I close the dishwasher and turn it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['toilet', 'detergent', 'ceilingfan', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'remote_control', 'envelope', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'mousepad', 'couch', 'wall', 'love_seat', 'newspaper', 'keys', 'television', 'toothbrush_holder', 'curtain', 'after_shave', 'clothes_scarf', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I'll watch TV", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'milk', 'knife', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'purse', 'computer', 'mat', 'tvstand', 'foundation', 'window', 'mousepad', 'closetdrawer', 'couch', 'bowl', 'wall', 'powersocket', 'keys', 'coffee_filter', 'mouse', 'dresser', 'home_office', 'light', 'coffee']\nTask:Hang keys\nresponse:\n{\n    \"thoughts\": Remove keys from purse, pocket or hand.  Place in designated area (bowl, or on hook). Leave area,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <purse>', '[GRAB] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[FIND] <bowl>', '[PUTBACK] <keys> <bowl>', '[CLOSE] <purse>', '[PUTOBJBACK] <purse>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'dog', 'toothbrush', 'bills', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'purse', 'faucet', 'balanceball', 'food_noodles', 'wall', 'bench', 'cup', 'photoframe', 'mouthwash', 'console', 'kitchen_counter', 'cupboard', 'sink', 'bag', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "go to bathroom. walk to sink,turn on faucet, put hands in water. rub hands together. grab soap. rub soap in hands. put soap down. put hands under water. rubs hands together. turn off faucet", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'walllamp', 'detergent', 'orchid', 'floor', 'ceiling', 'keyboard', 'remote_control', 'computer', 'button', 'alcohol', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'check', 'mouthwash', 'television', 'coffe_maker', 'curtain', 'dresser', 'clothes_dress', 'door', 'sink', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'hanger', 'floor', 'phone', 'desk', 'bookshelf', 'computer', 'piano_bench', 'stovefan', 'plate', 'tvstand', 'window', 'mousepad', 'wall', 'powersocket', 'box', 'photoframe', 'novel', 'doorjamb', 'towel_rack', 'dresser', 'cupboard', 'door', 'light', 'vase', 'colander', 'shoe_rack']\nTask:Read\nresponse:\n", "output": {"thoughts": "We pick out a book from the bookshelf and read it.", "actions": ["[WALK] <novel>", "[WALK] <bookshelf>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'nail_polish', 'knife', 'cpuscreen', 'floor', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'mat', 'mousepad', 'closetdrawer', 'couch', 'dining_room', 'wall', 'basket_for_clothes', 'cup', 'check', 'television', 'curtain', 'mouse', 'chessboard', 'shoes', 'comb', 'cupboard', 'hairdryer', 'door', 'home_office']\nTask:Read book\nresponse:\n{\n    \"thoughts\": find a book to read. sit down. read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['juice', 'freezer', 'ground_coffee', 'detergent', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'faucet', 'shower', 'balanceball', 'food_noodles', 'candle', 'bowl', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'toothbrush_holder', 'rag', 'doorjamb', 'towel_rack', 'conditioner', 'sink', 'bathroom_cabinet', 'soap', 'laser_pointer']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'walllamp', 'bathroom', 'trashcan', 'food_carrot', 'clothes_underwear', 'floor', 'desk', 'toaster', 'plate', 'shower', 'tvstand', 'food_noodles', 'clothes_socks', 'mousepad', 'wall', 'oven', 'cup', 'newspaper', 'coffe_maker', 'toilet_paper', 'food_donut', 'water', 'tray', 'food_onion', 'lighter', 'maindoor', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": first i will take all the dishes to the dish washer and with the help of dishwasher and water i will clean all the dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <water>', '[TURNTO] <water>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'bathroom', 'trashcan', 'ceilingfan', 'bills', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'folder', 'stovefan', 'tvstand', 'window', 'homework', 'wall', 'oven', 'powersocket', 'cutting_board', 'basket_for_clothes', 'food_cheese', 'clothes_gloves', 'check', 'television', 'toothbrush_holder', 'curtain', 'bathroom_cabinet', 'laser_pointer']\nTask:Open window\nresponse:\n", "output": {"thoughts": "Unlock the latch at the top. Push the bottom half of the window upwards.", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'dog', 'cat', 'drawing', 'food_egg', 'floor', 'phone', 'piano_bench', 'mat', 'mousepad', 'oil', 'food_jam', 'pillow', 'wall_clock', 'photoframe', 'coffe_maker', 'food_donut', 'food_rice', 'tray', 'after_shave', 'kitchen_counter', 'clothes_dress', 'door', 'sink', 'bag', 'bathroom_cabinet', 'table', 'dishwasher', 'creditcard']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to living room , find phone ,pick up phone 3,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'crayon', 'freezer', 'walllamp', 'bathroom', 'knifeblock', 'orchid', 'drawing', 'cpuscreen', 'longboard', 'bathroom_counter', 'character', 'bedroom', 'bookshelf', 'toaster', 'mat', 'foundation', 'nightstand', 'wall', 'oven', 'cup', 'television', 'tray', 'mouse', 'shoes', 'cupboard', 'bathroom_cabinet', 'light', 'tooth_paste']\nTask:Walk through\nresponse:\n", "output": {"thoughts": "Walk through the hallway.", "actions": ["[WALK] <home_office>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'mousepad', 'couch', 'food_kiwi', 'wall', 'bench', 'cup', 'check', 'television', 'curtain', 'napkin', 'table', 'soap', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch. I pick up the remote control. I push the power button.  I push the guide button and look for my favorite show \"The Middle.\"  I click that channel and enjoy my show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUSH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'freezer', 'ceilingfan', 'food_butter', 'drawing', 'orchid', 'food_egg', 'floor', 'ceiling', 'phone', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'deck_of_cards', 'window', 'mousepad', 'couch', 'wall', 'pasta', 'television', 'curtain', 'rag', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'door', 'table', 'creditcard']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "In the evenings after dinner, we sit on the couch, turn on the television, cue up Netflix and watch our favorite shows.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'toilet', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'phone', 'food_cake', 'computer', 'ceilinglamp', 'shoe-shine_kit', 'iron', 'stereo', 'electric_shaver', 'homework', 'balanceball', 'picture', 'window', 'dining_room', 'powersocket', 'cup', 'photoframe', 'mirror', 'tray', 'pencil', 'shoes', 'conditioner', 'sink', 'creditcard']\nTask:Shave\nresponse:\n{\n    \"thoughts\": Pick up razor.  Push the button to turn it on.  Move it in a linear motion against the beard growth.  Make sure to get the bottom of your chin, your cheeks, and mustache area.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <electric_shaver>', '[FIND] <electric_shaver>', '[GRAB] <electric_shaver>', '[FIND] <mirror>', '[TURNTO] <mirror>', '[SWITCHON] <electric_shaver>', '[SWITCHOFF] <electric_shaver>', '[PUTOBJBACK] <electric_shaver>']\n}\n", "input": "\nScene_Objects:['sheets', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'computer', 'food_bacon', 'stovefan', 'mat', 'tvstand', 'window', 'mousepad', 'form', 'dining_room', 'wall', 'cutting_board', 'bed', 'clothes_gloves', 'check', 'television', 'oven_mitts', 'curtain', 'mouse', 'hair', 'napkin', 'vase', 'instrument_violin', 'laser_pointer']\nTask:Gaze out window\nresponse:\n", "output": {"thoughts": "Look at the yard from the window", "actions": ["[WALK] <home_office>", "[WALK] <window>", "[FIND] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'ceilingfan', 'drawing', 'food_egg', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'television', 'food_rice', 'curtain', 'kitchen_counter', 'hair', 'home_office', 'table', 'light', 'dishwasher', 'dirt']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['juice', 'crayon', 'dog', 'cat', 'trashcan', 'ceilingfan', 'drawing', 'longboard', 'floor', 'ceiling', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'faucet', 'couch', 'food_peanut_butter', 'wall', 'keys', 'toilet_paper', 'television', 'curtain', 'towel_rack', 'mouse', 'shoes', 'vacuum_cleaner', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on the couch and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'trashcan', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'ceilinglamp', 'mat', 'faucet', 'shower', 'couch', 'towel', 'wall', 'bench', 'oven', 'photoframe', 'bed', 'mouse', 'kitchen_counter', 'dresser', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'drawing', 'hanger', 'floor', 'phone', 'ceilinglamp', 'plate', 'iron', 'candle', 'pillow', 'wall', 'bench', 'powersocket', 'oven', 'food_food', 'video_game_controller', 'check', 'wooden_spoon', 'after_shave', 'doorjamb', 'shoes', 'dresser', 'cupboard', 'clothes_dress', 'door', 'conditioner', 'sink', 'table', 'light']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "Put food on the table, sit down, and eat the food.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <plate> <table>", "[FIND] <chair>", "[SIT] <chair>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'food_pizza', 'dry_pasta', 'walllamp', 'trashcan', 'food_carrot', 'dough', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'wall', 'wall_clock', 'oven', 'food_cheese', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher']\nTask:Keep an eye on stove as something is cooking\nresponse:\n{\n    \"thoughts\": Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <pot>', '[TURNTO] <pot>', '[LOOKAT] <pot>', '[FIND] <wall_clock>', '[TURNTO] <wall_clock>', '[LOOKAT] <wall_clock>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'freezer', 'detergent', 'trashcan', 'drawing', 'longboard', 'floor', 'phone', 'purse', 'computer', 'bookmark', 'brush', 'window', 'wall', 'child', 'powersocket', 'bed', 'novel', 'food_food', 'after_shave', 'curtain', 'doorjamb', 'towel_rack', 'shoes', 'laundry_detergent', 'door', 'napkin', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Read story to children", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <child>", "[GREET] <child>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[CLOSE] <filing_cabinet>", "[WALK] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'foundation', 'homework', 'closetdrawer', 'couch', 'pillow', 'broom', 'nightstand', 'wall', 'oven', 'light_bulb', 'television', 'curtain', 'dresser', 'bathroom_cabinet', 'table', 'dishwasher', 'creditcard']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['fork', 'tea', 'chair', 'freezer', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'bedroom', 'purse', 'face', 'mop_bucket', 'plate', 'faucet', 'balanceball', 'tvstand', 'mousepad', 'dustpan', 'wall', 'face_soap', 'photoframe', 'curtain', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher', 'teeth']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Run warm water. Cleanse my face with soap. Rinse my face with water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <face_soap>", "[GRAB] <face_soap>", "[FIND] <face>", "[SCRUB] <face>", "[PUTOBJBACK] <face_soap>", "[WASH] <face>", "[RINSE] <face>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'measuring_cup', 'microwave', 'freezer', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'knifeblock', 'phone', 'bedroom', 'desk', 'button', 'stereo', 'faucet', 'picture', 'pillow', 'dining_room', 'nightstand', 'wall', 'bed', 'check', 'toothbrush_holder', 'tray', 'dresser', 'sink', 'home_office', 'table', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the radio; switch on the radio and listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['juice', 'food_salt', 'beer', 'microwave', 'trashcan', 'food_carrot', 'milk', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'ceilinglamp', 'tvstand', 'balanceball', 'deck_of_cards', 'window', 'mousepad', 'couch', 'wall', 'bench', 'television', 'curtain', 'mouse', 'controller', 'clothes_scarf', 'napkin', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on the couch and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'cat', 'trashcan', 'orchid', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'shower', 'window', 'mousepad', 'microphone', 'dining_room', 'wall', 'pasta', 'bench', 'oven', 'newspaper', 'check', 'television', 'tray', 'curtain', 'towel_rack', 'mouse', 'home_office', 'bathroom_cabinet']\nTask:Read newspaper\nresponse:\n{\n    \"thoughts\": Pull out dining room chair. Sit in dining room chair. Open newspaper to section you'd like to read. Read section. Keep reading all sections you'd like.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <newspaper>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'blender', 'freezer', 'cpuscreen', 'hanger', 'longboard', 'ceiling', 'desk', 'bedroom', 'toaster', 'faucet', 'pasta', 'bench', 'love_seat', 'food_cheese', 'newspaper', 'video_game_controller', 'wine', 'curtain', 'doorjamb', 'shaving_cream', 'kettle', 'shoes', 'hairdryer', 'clothes_scarf', 'napkin', 'spoon', 'teeth']\nTask:Read\nresponse:\n", "output": {"thoughts": "Reading a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'orchid', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bench', 'food_food', 'water', 'doorjamb', 'towel_rack', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I turn on faucet. I hold out my hands. I get my hands wet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <hands_both>', '[POUR] <water> <hands_both>']\n}\n", "input": "\nScene_Objects:['walllamp', 'bathroom', 'detergent', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'desk', 'thread', 'faucet', 'picture', 'mousepad', 'couch', 'wall', 'bed', 'photoframe', 'razor', 'doorjamb', 'console', 'mouse', 'cards', 'kitchen_counter', 'dresser', 'cupboard', 'hair', 'conditioner', 'sink', 'bag', 'vase', 'coffee']\nTask:Wipe down sink\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'toilet', 'food_butter', 'orchid', 'bathroom_counter', 'bedroom', 'tea_bag', 'bookshelf', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'window', 'food_noodles', 'couch', 'clothes_hat', 'pillow', 'bench', 'powersocket', 'basket_for_clothes', 'television', 'oven_mitts', 'kitchen_counter', 'shoes', 'pajamas', 'door', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sponge', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'remote_control', 'purse', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'oven', 'mouthwash', 'television', 'wine', 'curtain', 'after_shave', 'towel_rack', 'maindoor', 'clothes_scarf', 'hairdryer', 'table', 'light', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'keyboard', 'thread', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'mousepad', 'couch', 'wall', 'photoframe', 'bed', 'toilet_paper', 'food_onion', 'towel_rack', 'mouse', 'kitchen_counter', 'cd', 'hairbrush', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'walllamp', 'cat', 'ceilingfan', 'drawing', 'cpuscreen', 'clothes_underwear', 'longboard', 'floor', 'ceiling', 'keyboard', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'oil', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'ice', 'curtain', 'after_shave', 'door', 'table', 'dishwasher', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on couch and watch the tv", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'oil', 'couch', 'wall', 'cup', 'food_food', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'console', 'hairbrush', 'cupboard', 'food_steak', 'maindoor', 'door', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch favorite shows, netflix, hulu,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'blender', 'freezer', 'walllamp', 'nail_polish', 'trashcan', 'knifeblock', 'orchid', 'hanger', 'floor', 'bedroom', 'keyboard', 'computer', 'stereo', 'window', 'oil', 'couch', 'pillow', 'wall', 'cup', 'bed', 'food_donut', 'curtain', 'centerpiece', 'console', 'shoes', 'dresser', 'maindoor', 'home_office', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Turn on radio, put on favorite channel, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'toaster', 'mat', 'plate', 'shower', 'iron', 'dish_soap', 'closetdrawer', 'pillow', 'wall', 'oven', 'basket_for_clothes', 'coffe_maker', 'check', 'tray', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'dishwasher', 'board_game', 'dirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>', '[GRAB] <plate>', '[WIPE] <plate>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[PUTOBJBACK] <dishrack>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'toilet', 'chair', 'jelly', 'cat', 'dough', 'cpuscreen', 'floor', 'desk', 'bedroom', 'keyboard', 'computer', 'laptop', 'mat', 'tvstand', 'mousepad', 'wall', 'nightstand', 'love_seat', 'cup', 'clothes_gloves', 'television', 'towel_rack', 'mouse', 'dresser', 'controller', 'bathroom_cabinet', 'electrical_outlet']\nTask:Write book\nresponse:\n", "output": {"thoughts": "Writing a book", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[FIND] <electrical_outlet>", "[PLUGIN] <laptop>", "[WALK] <love_seat>", "[SIT] <love_seat>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'table_cloth', 'tvstand', 'faucet', 'stereo', 'window', 'couch', 'form', 'wall', 'powersocket', 'light_bulb', 'razor', 'vacuum_cleaner', 'controller', 'door', 'home_office', 'light', 'dishwasher', 'bathroom_cabinet', 'colander']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk into the bath room and go to the toilet,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['sheets', 'chair', 'walllamp', 'detergent', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'bathroom_counter', 'phone', 'folder', 'man', 'keyboard', 'mop_bucket', 'tvstand', 'faucet', 'picture', 'plate', 'mousepad', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'check', 'food_onion', 'clothes_scarf', 'spoon', 'light', 'dishwasher']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "When someone knocks I n the door walk through the hallway and greet them.", "actions": ["[RUN] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'jelly', 'chair', 'phone', 'ceiling', 'desk', 'bedroom', 'table_cloth', 'bookmark', 'stovefan', 'mat', 'tvstand', 'stereo', 'shoe-shine_kit', 'faucet', 'toaster', 'mousepad', 'powersocket', 'cutting_board', 'shelf', 'food_cheese', 'television', 'wooden_spoon', 'after_shave', 'food_onion', 'doorjamb', 'shoes', 'dishrack', 'vase', 'electrical_outlet']\nTask:Turn on radio\nresponse:\n{\n    \"thoughts\": Go to the radio.  Switch on the power button.  Tune the station.  Set the volume.  Listen to the music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[FIND] <electrical_outlet>', '[TURNTO] <electrical_outlet>', '[PLUGIN] <stereo>', '[SWITCHON] <electrical_outlet>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'shampoo', 'bookshelf', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'couch', 'pillow', 'wall', 'pasta', 'tape', 'box', 'coffee_table', 'keys', 'food_food', 'razor', 'television', 'toothbrush_holder', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'table', 'tooth_paste']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <coffee_table>", "[PUTBACK] <remote_control> <coffee_table>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'microwave', 'bathroom', 'trashcan', 'toothbrush', 'drawing', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'purse', 'food_bacon', 'ceilinglamp', 'plate', 'laptop', 'deck_of_cards', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'spoon', 'home_office', 'table', 'teeth']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'ceilingfan', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'face', 'mail', 'table_cloth', 'faucet', 'food_jam', 'towel', 'dining_room', 'wall', 'face_soap', 'toy', 'photoframe', 'keys', 'water', 'rag', 'pencil', 'shoes', 'dresser', 'sink', 'bathroom_cabinet', 'home_office', 'colander', 'hands_both']\nTask:Wash face\nresponse:\n", "output": {"thoughts": "Turn on water faucet. Get rag wet. Get face soap out. Squirt face soap in hands and put it on face. Rub soap into face. Wash soap off of face. Dry face off on the hand towel. Hang out rag and hand towel.", "actions": ["[WALK] <bathroom>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <rag> <sink>", "[GRAB] <rag>", "[FIND] <face_soap>", "[GRAB] <face_soap>", "[POUR] <face_soap> <rag>", "[PUTOBJBACK] <face_soap>", "[FIND] <face>", "[SCRUB] <face>", "[PUTBACK] <rag> <sink>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>", "[RINSE] <face>", "[WASH] <rag>", "[GRAB] <rag>", "[PUTOBJBACK] <rag>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <face>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'juice', 'microwave', 'chair', 'freezer', 'detergent', 'drawing', 'orchid', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'shower', 'mat', 'ceilinglamp', 'couch', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'newspaper', 'coffe_maker', 'television', 'ice', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'cupboard', 'table', 'dishwasher', 'creditcard']\nTask:Grab some juice\nresponse:\n{\n    \"thoughts\": I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_pizza', 'tablelamp', 'food_ice_cream', 'crayon', 'shampoo', 'keyboard', 'mail', 'bookmark', 'stovefan', 'faucet', 'pillow', 'dining_room', 'broom', 'wall', 'love_seat', 'photoframe', 'novel', 'food_food', 'wine', 'tray', 'after_shave', 'shaving_cream', 'kettle', 'headset', 'pajamas', 'laundry_detergent', 'door', 'bathroom_cabinet', 'shoe_rack']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[READ] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[PUTBACK] <bookmark> <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'walllamp', 'ground_coffee', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'stovefan', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'towel', 'wall', 'wall_clock', 'oven', 'television', 'tray', 'rag', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <towel>', '[GRAB] <towel>', '[CLOSE] <bathroom_cabinet>', '[FIND] <shower>', '[RINSE] <shower>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'microwave', 'toilet', 'freezer', 'dough', 'food_carrot', 'drawing', 'knife', 'clothes_underwear', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'keyboard', 'table_cloth', 'ceilinglamp', 'plate', 'couch', 'pillow', 'bowl', 'wall', 'nightstand', 'bench', 'coffe_maker', 'toothbrush_holder', 'diary', 'laundry_detergent', 'table', 'coffee']\nTask:Cut bread\nresponse:\n", "output": {"thoughts": "I get a slice of bread, grab a knife, and then cut the piece of bread in half.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <kitchen_cabinet>", "[FIND] <table>", "[PUTBACK] <plate> <table>", "[PUTBACK] <knife> <table>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[PUTBACK] <food_bread> <plate>", "[GRAB] <knife>", "[CUT] <food_bread>", "[PUTOBJBACK] <knife>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'bathroom', 'drawing', 'knifeblock', 'floor', 'character', 'keyboard', 'stovefan', 'mat', 'ceilinglamp', 'stereo', 'shower', 'window', 'toaster', 'wall', 'nightstand', 'box', 'food_cheese', 'bed', 'after_shave', 'lighter', 'diary', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'food_steak', 'light', 'dishwasher', 'clothes_shirt']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Turn on radio, put on favorite channel, listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'freezer', 'walllamp', 'knifeblock', 'drawing', 'food_egg', 'floor', 'ceiling', 'stovefan', 'ceilinglamp', 'laptop', 'faucet', 'oil', 'wall', 'bench', 'box', 'cup', 'food_food', 'check', 'coffee_filter', 'lighter', 'centerpiece', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'spoon', 'table', 'colander', 'board_game']\nTask:Grab some juice\nresponse:\n", "output": {"thoughts": "I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTOBJBACK] <juice>", "[CLOSE] <freezer>", "[FIND] <table>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'toaster', 'stovefan', 'plate', 'faucet', 'iron', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'food_food', 'wooden_spoon', 'tray', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'filing_cabinet', 'microwave', 'walllamp', 'ceilingfan', 'drawing', 'clothes_underwear', 'bathroom_counter', 'dvd_player', 'character', 'stamp', 'keyboard', 'food_cake', 'shower', 'picture', 'window', 'electric_shaver', 'couch', 'broom', 'box', 'food_food', 'toothbrush_holder', 'food_donut', 'after_shave', 'food_onion', 'mouse', 'hairdryer', 'sink', 'table', 'clothes_shirt']\nTask:Open bathroom window\nresponse:\n", "output": {"thoughts": "Make sure bathroom window is unlocked. Lift up bathroom window so it is open.", "actions": ["[WALK] <bathroom>", "[WALK] <window>", "[FIND] <window>", "[OPEN] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'mop', 'bathroom', 'food_egg', 'clothes_underwear', 'bathroom_counter', 'phone', 'ceiling', 'computer', 'toaster', 'mat', 'shoe-shine_kit', 'shower', 'clothes_hat', 'clothes_pants', 'food_orange', 'food_kiwi', 'basket_for_clothes', 'coffe_maker', 'television', 'curtain', 'mouse', 'cd', 'comb', 'clothes_dress', 'clothes_skirt', 'soap', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I need to get the dirty clothes basket.  Walk to bathroom.  Put all the clothes in washing machine.  Add soap.  Turn on machine.,\n    \"actions\": ['[WALK] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'walllamp', 'bathroom', 'ceilingfan', 'trashcan', 'drawing', 'floor', 'ceiling', 'food_bacon', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'couch', 'wall', 'nightstand', 'wall_clock', 'bench', 'bed', 'food_food', 'television', 'coin', 'curtain', 'mouse', 'controller', 'table', 'vase', 'instrument_violin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sitting on the couch watching my favorite television show.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'ceiling', 'bookshelf', 'envelope', 'toaster', 'mat', 'food_orange', 'pillow', 'wall', 'wall_clock', 'powersocket', 'cutting_board', 'bed', 'oven_mitts', 'coin', 'doorjamb', 'lighter', 'pencil', 'kitchen_counter', 'dresser', 'pajamas', 'door', 'home_office', 'light', 'colander']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk towards to the bedroom, and I open the door of the bedroom, and I switch on the lights inside the bedroom, the switch is beside the door.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'measuring_cup', 'toilet', 'freezer', 'walllamp', 'food_carrot', 'knife', 'bathroom_counter', 'stamp', 'food_bacon', 'tvstand', 'plate', 'couch', 'oven', 'cup', 'photoframe', 'check', 'television', 'coffee_filter', 'toothbrush_holder', 'food_rice', 'tray', 'kitchen_counter', 'shoes', 'hairbrush', 'napkin', 'spoon', 'table', 'vase', 'creditcard']\nTask:Shut off alarm\nresponse:\n", "output": {"thoughts": "Shutting of the alarm clock", "actions": ["[WALK] <bedroom>", "[WALK] <alarm_clock>", "[FIND] <alarm_clock>", "[GRAB] <alarm_clock>", "[SWITCHOFF] <alarm_clock>", "[PUTOBJBACK] <alarm_clock>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'ground_coffee', 'bathroom', 'detergent', 'bathroom_counter', 'bedroom', 'bookshelf', 'button', 'toaster', 'mat', 'ceilinglamp', 'window', 'mousepad', 'couch', 'dining_room', 'photoframe', 'bed', 'novel', 'coffe_maker', 'food_food', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'sink', 'dishwasher', 'clothes_skirt', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to living room , find phone ,pick up phone 3,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'walllamp', 'detergent', 'ceilingfan', 'dough', 'milk', 'knifeblock', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'stovefan', 'ceilinglamp', 'faucet', 'iron', 'wall', 'nightstand', 'cup', 'food_food', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'food_steak', 'sink', 'bag', 'dishwasher', 'vase']\nTask:Get something to drink\nresponse:\n", "output": {"thoughts": "Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[CLOSE] <freezer>", "[POUR] <milk> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'stovefan', 'mat', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'pasta', 'powersocket', 'television', 'tray', 'curtain', 'doorjamb', 'mouse', 'dresser', 'napkin', 'bathroom_cabinet', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I want to relax on the sofa to some relaxing music in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <stereo>', '[PLUGIN] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'dustpan', 'dining_room', 'wall', 'nightstand', 'pasta', 'tape', 'photoframe', 'tray', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'bathroom_cabinet', 'light']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Press the power button on the computer to turn it on", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'stovefan', 'window', 'mat', 'plate', 'couch', 'candle', 'microphone', 'dining_room', 'wall', 'powersocket', 'box', 'photoframe', 'coffe_maker', 'television', 'curtain', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on the TV, sit on the couch, find a show, and watch it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['dog', 'food_egg', 'bathroom_counter', 'folder', 'bedroom', 'computer', 'table_cloth', 'piano_bench', 'stovefan', 'tvstand', 'faucet', 'foundation', 'homework', 'couch', 'nightstand', 'powersocket', 'oven', 'basket_for_clothes', 'cup', 'keys', 'food_food', 'mouthwash', 'after_shave', 'doorjamb', 'shoes', 'controller', 'maindoor', 'bag', 'dirt', 'coffee']\nTask:Get ready to leave\nresponse:\n", "output": {"thoughts": "Get your jacket and shoes on, locate the keys off the side table, walk out the front door and secure building with keys when you leave", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>", "[FIND] <keys>", "[GRAB] <keys>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'video_game_console', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'candle', 'broom', 'wall', 'powersocket', 'check', 'video_game_controller', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'hairbrush', 'door', 'table', 'dishwasher', 'dirt']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'jelly', 'freezer', 'food_carrot', 'drawing', 'needle', 'knifeblock', 'longboard', 'floor', 'character', 'ceilinglamp', 'shower', 'window', 'foundation', 'couch', 'wall', 'nightstand', 'food_cheese', 'newspaper', 'coffee_table', 'food_food', 'check', 'television', 'ice', 'doorjamb', 'kettle', 'maindoor', 'table', 'vase', 'board_game']\nTask:Read magazine\nresponse:\n", "output": {"thoughts": "Get magazine off coffee table. Sit in arm chair. Open magazine to first page. Read until finished.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <newspaper>", "[PUTBACK] <newspaper> <coffee_table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'blender', 'toilet', 'detergent', 'knife', 'hanger', 'floor', 'character', 'food_cake', 'brush', 'mat', 'plate', 'faucet', 'dish_soap', 'oil', 'closetdrawer', 'food_jam', 'dining_room', 'food_kiwi', 'oven', 'powersocket', 'cup', 'after_shave', 'kitchen_counter', 'dishrack', 'pajamas', 'spoon', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find dishwasher. Open dishwasher. Find dishrack. Find plate 1. Grab plate 1. Put plate 1 in dishrack. Repeat for dish 2. Find dish soap. Grab dish soap. Put dish soap to dishwasher. Close dishwasher. Switch on dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'shampoo', 'ceiling', 'folder', 'face', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'picture', 'oil', 'food_orange', 'dining_room', 'bench', 'box', 'newspaper', 'check', 'video_game_controller', 'mirror', 'curtain', 'towel_rack', 'hair', 'clothes_dress', 'laundry_detergent', 'bathroom_cabinet', 'table', 'clothes_skirt', 'vase', 'scrabble']\nTask:Check appearance in mirror\nresponse:\n", "output": {"thoughts": "Walk to full-length mirror. Look in mirror. Make sure everything about your appearance is adequate. Adjust clothes as needed.", "actions": ["[WALK] <bedroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>", "[FIND] <face>", "[TURNTO] <face>", "[LOOKAT] <face>", "[FIND] <hair>", "[TURNTO] <hair>", "[LOOKAT] <hair>", "[FIND] <clothes_dress>", "[TURNTO] <clothes_dress>", "[LOOKAT] <clothes_dress>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'filing_cabinet', 'microwave', 'chair', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'face', 'toaster', 'stovefan', 'faucet', 'oil', 'towel', 'wall', 'facial_cleanser', 'check', 'coffe_maker', 'tray', 'rag', 'mouse', 'kitchen_counter', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap']\nTask:Wash face\nresponse:\n{\n    \"thoughts\": Turn the faucet on. Wet the face. Take facial cleanser out of packaging. Lather on face. Rinse face with water. Turn off the faucet. Dry face with towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <face>', '[WASH] <face>', '[FIND] <rag>', '[GRAB] <rag>', '[RINSE] <rag>', '[FIND] <facial_cleanser>', '[GRAB] <facial_cleanser>', '[POUR] <facial_cleanser> <rag>', '[PUTOBJBACK] <facial_cleanser>', '[SCRUB] <face>', '[RINSE] <rag>', '[PUTOBJBACK] <rag>', '[RINSE] <face>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <face>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['trashcan', 'dough', 'drawing', 'needle', 'longboard', 'floor', 'desk', 'bedroom', 'computer', 'toaster', 'ceilinglamp', 'stovefan', 'shower', 'window', 'floor_lamp', 'mousepad', 'wall', 'nightstand', 'bed', 'television', 'coffee_filter', 'curtain', 'rag', 'doorjamb', 'cards', 'mouse', 'pencil', 'maindoor', 'sink', 'door']\nTask:Turn on lights\nresponse:\n", "output": {"thoughts": "Walk to your lamp, switch it on, and lay on your bed", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'walllamp', 'blow_dryer', 'knifeblock', 'hanger', 'floor', 'ceiling', 'piano_bench', 'ceilinglamp', 'faucet', 'stereo', 'oil', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'coffe_maker', 'toilet_paper', 'coffee_filter', 'water', 'wine', 'tray', 'curtain', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'dishwasher', 'laser_pointer', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Go to the kitchen and make coffee.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[PUTOBJBACK] <coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'bathroom', 'trashcan', 'food_butter', 'food_carrot', 'floor', 'dvd_player', 'desk', 'food_cake', 'button', 'mat', 'shower', 'stereo', 'clothes_socks', 'couch', 'oven', 'food_food', 'television', 'tray', 'food_onion', 'coffee_pot', 'pencil', 'kitchen_counter', 'cd', 'cupboard', 'clothes_dress', 'food_steak', 'electrical_outlet', 'creditcard']\nTask:Turn on radio\nresponse:\n", "output": {"thoughts": "Go to the radio.  Switch on the power button.  Tune the station.  Set the volume.  Listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <electrical_outlet>", "[TURNTO] <electrical_outlet>", "[PLUGIN] <stereo>", "[SWITCHON] <electrical_outlet>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'toilet', 'chair', 'freezer', 'detergent', 'food_butter', 'wallshelf', 'hanger', 'phone', 'character', 'desk', 'remote_control', 'button', 'piano_bench', 'toaster', 'mat', 'tvstand', 'picture', 'oil', 'couch', 'clothes_pants', 'pillow', 'towel', 'wall', 'novel', 'razor', 'diary', 'conditioner', 'table', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'beer', 'toilet', 'band-aids', 'nail_polish', 'ceilingfan', 'food_carrot', 'toothbrush', 'knife', 'shampoo', 'ceiling', 'bedroom', 'stamp', 'tea_bag', 'mop_bucket', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'homework', 'food_noodles', 'dining_room', 'bed', 'ice', 'mouse', 'spoon', 'home_office', 'bathroom_cabinet', 'vase', 'colander']\nTask:Shut front door\nresponse:\n", "output": {"thoughts": "Close the storm door. Lock the storm door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'nail_polish', 'cat', 'drawing', 'knife', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'photoframe', 'toilet_paper', 'check', 'television', 'doorjamb', 'towel_rack', 'mouse', 'vacuum_cleaner', 'dresser', 'cupboard', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'bathroom', 'trashcan', 'ceilingfan', 'knifeblock', 'clothes_underwear', 'hanger', 'longboard', 'floor', 'phone', 'table_cloth', 'ceilinglamp', 'balanceball', 'window', 'couch', 'dustpan', 'powersocket', 'bench', 'cup', 'newspaper', 'photoframe', 'coin', 'console', 'mouse', 'comb', 'maindoor', 'spoon', 'light', 'shoe_rack']\nTask:Leave home\nresponse:\n", "output": {"thoughts": "Put on shoes. Put on coat. Open door and exit through it. Lock door behind you.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'band-aids', 'trashcan', 'dough', 'food_egg', 'cpuscreen', 'floor', 'ceiling', 'folder', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'pillow', 'wall', 'powersocket', 'newspaper', 'food_food', 'check', 'oven_mitts', 'food_donut', 'food_onion', 'doorjamb', 'kettle', 'mouse', 'pajamas', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": go to the kitchen and put the groceries in the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'fork', 'crayon', 'tea', 'toilet', 'bathroom', 'orchid', 'hanger', 'floor', 'character', 'bedroom', 'tea_bag', 'keyboard', 'purse', 'toaster', 'mat', 'deck_of_cards', 'window', 'food_orange', 'pillow', 'photoframe', 'food_food', 'coffe_maker', 'food_donut', 'towel_rack', 'mouse', 'pencil', 'dresser', 'door', 'sink']\nTask:Open door\nresponse:\n", "output": {"thoughts": "Open the door when leaving or coming in.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'measuring_cup', 'chair', 'trashcan', 'toothbrush', 'cpuscreen', 'hanger', 'ceiling', 'character', 'keyboard', 'alcohol', 'mat', 'closetdrawer', 'form', 'dining_room', 'nightstand', 'oven', 'bench', 'bed', 'photoframe', 'food_food', 'coffe_maker', 'toothbrush_holder', 'tray', 'towel_rack', 'kitchen_counter', 'bathtub', 'sink', 'home_office', 'tooth_paste']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>']\n}\n", "input": "\nScene_Objects:['chair', 'bills', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'table_cloth', 'oil', 'mousepad', 'couch', 'microphone', 'dining_room', 'wall', 'coffe_maker', 'coin', 'tray', 'centerpiece', 'console', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'vase', 'teeth']\nTask:Put in chair\nresponse:\n", "output": {"thoughts": "Take a comfortable chair. Place it in the entrance hall", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[GRAB] <chair>", "[WALK] <home_office>", "[WALK] <floor>", "[FIND] <floor>", "[PUTBACK] <chair> <floor>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'cat', 'drawing', 'wallshelf', 'floor', 'ceiling', 'character', 'desk', 'bedroom', 'bookshelf', 'computer', 'food_bacon', 'bookmark', 'brush', 'mat', 'mousepad', 'clothes_hat', 'wall', 'nightstand', 'powersocket', 'wall_clock', 'cup', 'oven_mitts', 'doorjamb', 'door', 'sink', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": push any ligth button to turn it on, for example in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'crayon', 'microwave', 'nail_polish', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'couch', 'dining_room', 'wall', 'nightstand', 'oven', 'novel', 'television', 'curtain', 'dresser', 'pajamas', 'table', 'instrument_violin']\nTask:Turn off TV\nresponse:\n", "output": {"thoughts": "Walk over to tv and press power button", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'toothbrush', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'thread', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'plate', 'dining_room', 'wall', 'wall_clock', 'oven', 'toothbrush_holder', 'tray', 'doorjamb', 'towel_rack', 'chessboard', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": getting ready for bed,\n    \"actions\": ['[WALK] <bathroom_counter>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <toothbrush> <bathroom_counter>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>', '[GRAB] <toothbrush>', '[SCRUB] <toothbrush>', '[RINSE] <toothbrush>', '[FIND] <toothbrush_holder>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'dog', 'ceilingfan', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'table_cloth', 'stovefan', 'electric_shaver', 'dining_room', 'wall', 'nightstand', 'bench', 'powersocket', 'coffe_maker', 'ice', 'lighter', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'home_office', 'table', 'light', 'vase']\nTask:Push all chairs in\nresponse:\n", "output": {"thoughts": "Go around the dining room table, going to each chair, and push all the chairs in neatly.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <chair>", "[PUSH] <chair>", "[FIND] <chair>", "[PUSH] <chair>", "[FIND] <chair>", "[PUSH] <chair>", "[FIND] <chair>", "[PUSH] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'detergent', 'orchid', 'bathroom_counter', 'character', 'keyboard', 'remote_control', 'bookshelf', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'dustpan', 'nightstand', 'bench', 'tape', 'cup', 'bed', 'coffee_table', 'wooden_spoon', 'doorjamb', 'towel_rack', 'pencil', 'kitchen_counter', 'cd', 'dresser', 'hair']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <coffee_table>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'freezer', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'mop_bucket', 'shower', 'homework', 'candle', 'clothes_pants', 'pillow', 'wall', 'oven', 'bed', 'keys', 'check', 'coffe_maker', 'tray', 'centerpiece', 'pajamas', 'bag', 'spoon', 'soap', 'laser_pointer', 'scrabble']\nTask:Put away keys\nresponse:\n", "output": {"thoughts": "Return keys into a bag", "actions": ["[WALK] <home_office>", "[WALK] <keys>", "[FIND] <keys>", "[GRAB] <keys>", "[FIND] <bag>", "[OPEN] <bag>", "[PUTBACK] <keys> <bag>", "[CLOSE] <bag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'drawing', 'orchid', 'clothes_underwear', 'hanger', 'desk', 'clothes_jacket', 'mop_bucket', 'bookmark', 'shower', 'electric_shaver', 'clothes_socks', 'clothes_pants', 'food_kiwi', 'bench', 'basket_for_clothes', 'photoframe', 'bed', 'coffe_maker', 'food_rice', 'shaving_cream', 'dishrack', 'cupboard', 'clothes_dress', 'laundry_detergent', 'clothes_skirt', 'soap', 'board_game', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bathroom, take the dirt clothes and put into the washing machine, then put the soap, set the cleaning program and switch on the machine,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_skirt>', '[GRAB] <clothes_skirt>', '[PUTBACK] <clothes_skirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[CLOSE] <basket_for_clothes>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_ice_cream', 'food_salt', 'toilet', 'chair', 'detergent', 'orchid', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'phone', 'bookshelf', 'computer', 'clothes_jacket', 'piano_bench', 'pillow', 'wall', 'powersocket', 'photoframe', 'doorjamb', 'cd', 'dishrack', 'dresser', 'cupboard', 'door', 'home_office', 'light', 'bathroom_cabinet', 'table']\nTask:Sit in chair\nresponse:\n", "output": {"thoughts": "I sit in a chair to rest and eat.", "actions": ["[WALK] <dining_room>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'mat', 'ceilinglamp', 'tvstand', 'homework', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk to the office. I push the on button on the computer. I sit down in the chair. I work on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <button>', '[FIND] <button>', '[PUSH] <button>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['food_bread', 'food_ice_cream', 'food_salt', 'crayon', 'microwave', 'freezer', 'walllamp', 'milk', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'toaster', 'ceilinglamp', 'foundation', 'food_noodles', 'wall', 'oven', 'cup', 'bed', 'food_food', 'coffe_maker', 'toilet_paper', 'coin', 'tray', 'food_onion', 'kettle', 'light', 'dishwasher', 'electrical_outlet']\nTask:Toast bread\nresponse:\n", "output": {"thoughts": "Place two slices of bread in the toaster. Select medium toast on toaster.  Push down button to begin toasting.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[WALK] <electrical_outlet>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <electrical_outlet>", "[TURNTO] <toaster>", "[LOOKAT] <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'walllamp', 'bills', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bathroom_counter', 'computer', 'tvstand', 'form', 'towel', 'wall', 'oven', 'photoframe', 'clothes_gloves', 'bed', 'food_food', 'doorjamb', 'towel_rack', 'dishrack', 'dresser', 'cupboard', 'door', 'light', 'table', 'colander', 'instrument_violin']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk over to the phone and pick it up.,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'chair', 'jelly', 'walllamp', 'cat', 'trashcan', 'hanger', 'floor', 'ceiling', 'phone', 'food_bacon', 'toaster', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'couch', 'wall', 'oven', 'bed', 'tray', 'kitchen_counter', 'shoes', 'dishrack', 'dresser', 'cupboard', 'bathroom_cabinet', 'light', 'teeth']\nTask:Put away shoes\nresponse:\n", "output": {"thoughts": "Pick up shoes from entry hall. Open entry way closet. Put shoes in closet. Close closet door.", "actions": ["[WALK] <home_office>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[FIND] <dresser>", "[OPEN] <dresser>", "[PUTIN] <shoes> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'cpuscreen', 'phone', 'desk', 'cleaning_solution', 'keyboard', 'purse', 'table_cloth', 'stovefan', 'tvstand', 'shower', 'stereo', 'food_noodles', 'towel', 'pillow', 'wall', 'wall_clock', 'oven', 'basket_for_clothes', 'photoframe', 'novel', 'food_rice', 'rag', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'napkin', 'table']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[POUR] <cleaning_solution> <rag>', '[PUTOBJBACK] <cleaning_solution>', '[WALK] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[WIPE] <mirror>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['glue', 'mop', 'toilet', 'cat', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'orchid', 'floor', 'ceiling', 'computer', 'clothes_jacket', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'couch', 'wall', 'nightstand', 'food_food', 'mouthwash', 'television', 'food_donut', 'curtain', 'after_shave', 'dresser', 'table']\nTask:Look at painting\nresponse:\n", "output": {"thoughts": "Walk to the painting.  Look up at the painting.  Admire it on the wall.", "actions": ["[WALK] <home_office>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'ceilinglamp', 'floor_lamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'oven', 'photoframe', 'novel', 'television', 'doorjamb', 'dresser', 'cupboard', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'microwave', 'chair', 'freezer', 'nail_polish', 'bathroom', 'food_butter', 'knifeblock', 'cpuscreen', 'hanger', 'ceiling', 'dvd_player', 'stamp', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'iron', 'nightstand', 'oven', 'cutting_board', 'check', 'television', 'coin', 'tray', 'doorjamb', 'bathroom_cabinet', 'teeth']\nTask:Answer door\nresponse:\n", "output": {"thoughts": "Walk up to the door.  Open the door and greet whomever is out side", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'drawing', 'hanger', 'floor', 'ceiling', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'foundation', 'oil', 'couch', 'pillow', 'wall', 'wall_clock', 'cup', 'novel', 'television', 'curtain', 'after_shave', 'door', 'napkin', 'home_office', 'table']\nTask:Turn off TV\nresponse:\n{\n    \"thoughts\": Walk over to tv and press power button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'scissors', 'freezer', 'band-aids', 'blow_dryer', 'bathroom', 'ceilingfan', 'drawing', 'bathroom_counter', 'ceiling', 'food_cake', 'toaster', 'mat', 'faucet', 'shower', 'mousepad', 'food_jam', 'clothes_pants', 'bowl', 'wall', 'photoframe', 'keys', 'novel', 'toilet_paper', 'tray', 'controller', 'sink', 'bathroom_cabinet', 'table']\nTask:Get toilet paper\nresponse:\n", "output": {"thoughts": "There are extra toilet papers in the bottom cabinet.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <toilet_paper>", "[GRAB] <toilet_paper>", "[CLOSE] <bathroom_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'juice', 'scissors', 'beer', 'chair', 'freezer', 'bathroom', 'ceilingfan', 'hanger', 'floor', 'computer', 'food_bacon', 'mat', 'couch', 'creditcard', 'dining_room', 'oven', 'tape', 'bed', 'light_bulb', 'novel', 'food_food', 'mouse', 'cupboard', 'conditioner', 'home_office', 'table', 'scrabble', 'shoe_rack']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['juice', 'food_salt', 'walllamp', 'detergent', 'trashcan', 'knifeblock', 'drawing', 'floor', 'stamp', 'table_cloth', 'plate', 'faucet', 'deck_of_cards', 'clothes_pants', 'wall', 'bench', 'powersocket', 'tape', 'photoframe', 'food_food', 'tray', 'pencil', 'kitchen_counter', 'cupboard', 'maindoor', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'coffee']\nTask:Bring dirty plate to sink\nresponse:\n", "output": {"thoughts": "Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[TURNTO] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <sink>", "[FIND] <sink>", "[PUTBACK] <plate> <sink>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'chair', 'walllamp', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'love_seat', 'photoframe', 'television', 'doorjamb', 'dresser', 'door', 'clothes_scarf']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit and watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'trashcan', 'food_carrot', 'orchid', 'floor', 'ceiling', 'thread', 'mail', 'button', 'food_bacon', 'bookmark', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'wall', 'bench', 'cup', 'food_food', 'razor', 'television', 'food_donut', 'towel_rack', 'console', 'kitchen_counter', 'hairbrush', 'cupboard', 'sink', 'vase']\nTask:Added meat to freezer\nresponse:\n", "output": {"thoughts": "After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'detergent', 'orchid', 'knifeblock', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'bookshelf', 'mat', 'shower', 'faucet', 'wall', 'powersocket', 'oven', 'light_bulb', 'novel', 'check', 'television', 'food_donut', 'curtain', 'doorjamb', 'kitchen_counter', 'spoon', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Switch on light in home office,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[SWITCHON] <light>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'microwave', 'toilet', 'chair', 'walllamp', 'cpuscreen', 'hanger', 'phone', 'computer', 'stereo', 'faucet', 'window', 'couch', 'dining_room', 'wall_clock', 'bench', 'pasta', 'coffe_maker', 'food_food', 'doorjamb', 'mouse', 'pajamas', 'controller', 'sink', 'napkin', 'light', 'table', 'dishwasher']\nTask:Walk to room\nresponse:\n", "output": {"thoughts": "Leave the living room and enter the hall", "actions": ["[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'microphone', 'form', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'coffe_maker', 'television', 'curtain', 'vacuum_cleaner', 'dishrack', 'dresser', 'hairbrush', 'door', 'table', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'mop', 'toilet', 'freezer', 'walllamp', 'food_carrot', 'knife', 'ceiling', 'purse', 'computer', 'table_cloth', 'faucet', 'clothes_pants', 'bench', 'photoframe', 'check', 'food_food', 'television', 'toilet_paper', 'ice', 'curtain', 'rag', 'lighter', 'mouse', 'pencil', 'vacuum_cleaner', 'door', 'clothes_scarf', 'sink', 'clothes_skirt']\nTask:Look at mirror\nresponse:\n", "output": {"thoughts": "I look at my reflection to check my cleanliness.", "actions": ["[WALK] <bathroom>", "[WALK] <mirror>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[LOOKAT] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'walllamp', 'bathroom', 'detergent', 'knifeblock', 'bathroom_counter', 'purse', 'alcohol', 'bookmark', 'stovefan', 'toaster', 'stereo', 'food_jam', 'dining_room', 'powersocket', 'food_food', 'mouthwash', 'tray', 'after_shave', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office', 'laser_pointer']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the hallway.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dining_room>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['beer', 'crayon', 'freezer', 'detergent', 'orchid', 'drawing', 'floor', 'ceiling', 'toaster', 'ceilinglamp', 'faucet', 'window', 'electric_shaver', 'couch', 'clothes_pants', 'wall', 'photoframe', 'newspaper', 'food_food', 'television', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'shaving_cream', 'comb', 'hair', 'home_office', 'table', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a can of beer, open it, sit on couch, relax", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <beer>", "[GRAB] <beer>", "[CLOSE] <freezer>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[DRINK] <beer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'blender', 'toilet', 'bathroom', 'detergent', 'food_butter', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'mousepad', 'couch', 'food_kiwi', 'wall', 'oven', 'toy', 'coffe_maker', 'television', 'oven_mitts', 'coffee_filter', 'ice', 'coffee_pot', 'shaving_cream', 'kitchen_counter', 'cd', 'cupboard', 'sink', 'home_office', 'dishwasher', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[WALK] <coffee>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'beer', 'chair', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'dvd_player', 'stamp', 'bookshelf', 'toaster', 'ceilinglamp', 'faucet', 'homework', 'floor_lamp', 'window', 'food_orange', 'dining_room', 'nightstand', 'bench', 'food_cheese', 'television', 'doorjamb', 'lighter', 'cupboard', 'home_office', 'dishwasher', 'laser_pointer']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Walk to the lamp.  Find the switch.  Click the switch off.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[SWITCHOFF] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'tablelamp', 'juice', 'food_salt', 'microwave', 'chair', 'freezer', 'detergent', 'trashcan', 'toothbrush', 'floor', 'ceiling', 'bookshelf', 'toaster', 'shower', 'stereo', 'mousepad', 'clothes_pants', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'food_food', 'check', 'food_donut', 'kitchen_counter', 'cupboard', 'door', 'home_office']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Go kitchen and make something for eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>']\n}\n", "input": "\nScene_Objects:['beer', 'microwave', 'drawing', 'floor', 'purse', 'computer', 'plate', 'balanceball', 'shower', 'stereo', 'dustpan', 'clothes_hat', 'microphone', 'bench', 'basket_for_clothes', 'cup', 'bed', 'check', 'coffe_maker', 'food_onion', 'after_shave', 'cards', 'pencil', 'chessboard', 'maindoor', 'napkin', 'table', 'board_game', 'clothes_shirt', 'coffee']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <plate> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'walllamp', 'knifeblock', 'drawing', 'floor', 'phone', 'keyboard', 'toaster', 'stovefan', 'plate', 'shower', 'tvstand', 'dish_soap', 'couch', 'wall', 'coffe_maker', 'video_game_controller', 'food_onion', 'rag', 'towel_rack', 'mouse', 'kettle', 'kitchen_counter', 'cupboard', 'door', 'sink', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'toilet', 'detergent', 'orchid', 'longboard', 'floor', 'desk', 'food_bacon', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'balanceball', 'window', 'bed', 'oven_mitts', 'ice', 'curtain', 'after_shave', 'lighter', 'towel_rack', 'cards', 'controller', 'laundry_detergent', 'maindoor', 'door', 'sink', 'home_office', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Twist the lamp switch until the light goes on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'fork', 'pot', 'sheets', 'food_salt', 'detergent', 'food_butter', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'shampoo', 'desk', 'food_cake', 'mat', 'faucet', 'food_kiwi', 'wall', 'powersocket', 'toy', 'novel', 'food_food', 'curtain', 'dresser', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Going to the sink and scrubbing hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'walllamp', 'blow_dryer', 'toothbrush', 'drawing', 'kitchen_cabinet', 'floor', 'ceiling', 'character', 'bedroom', 'keyboard', 'ceilinglamp', 'stereo', 'shower', 'foundation', 'food_peanut_butter', 'towel', 'wall', 'clothes_gloves', 'food_food', 'toilet_paper', 'tray', 'after_shave', 'dresser', 'cupboard', 'door', 'light']\nTask:Look in refrigerator\nresponse:\n", "output": {"thoughts": "Approach the refrigerator. Use the handle to open the refrigerator door. Look at each shelf of the refrigerator. Look in each refrigerator drawer. Close the refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[WALK] <freezer>", "[OPEN] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <food_food>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'jelly', 'freezer', 'chair', 'blow_dryer', 'bathroom', 'walllamp', 'knife', 'phone', 'keyboard', 'stovefan', 'homework', 'couch', 'food_jam', 'bench', 'oven', 'toy', 'box', 'food_cheese', 'oven_mitts', 'food_rice', 'tray', 'curtain', 'after_shave', 'dresser', 'clothes_dress', 'conditioner', 'hairdryer', 'vase', 'board_game']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my living room, find my cat, scrub it using my hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'tea', 'mop', 'microwave', 'toilet', 'toothbrush', 'drawing', 'knifeblock', 'orchid', 'phone', 'ceiling', 'computer', 'stovefan', 'toaster', 'tvstand', 'window', 'floor_lamp', 'couch', 'clothes_hat', 'photoframe', 'after_shave', 'food_onion', 'headset', 'chessboard', 'cupboard', 'controller', 'door', 'bathroom_cabinet', 'clothes_skirt']\nTask:Switch on lamp\nresponse:\n", "output": {"thoughts": "Press the button so the light comes on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'beer', 'chair', 'freezer', 'knifeblock', 'bathroom_counter', 'longboard', 'phone', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'balanceball', 'stereo', 'mousepad', 'clothes_pants', 'nightstand', 'bench', 'bed', 'photoframe', 'novel', 'television', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'food_steak', 'sink', 'hairdryer', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the book shelf, get a book, open the book and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'chair', 'blow_dryer', 'trashcan', 'orchid', 'knifeblock', 'drawing', 'longboard', 'floor', 'ceiling', 'desk', 'cleaning_solution', 'computer', 'window', 'oil', 'clothes_socks', 'couch', 'pillow', 'wall', 'bed', 'newspaper', 'novel', 'toilet_paper', 'food_donut', 'curtain', 'rag', 'doorjamb', 'maindoor', 'napkin', 'bathroom_cabinet']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Gather wood cleaner and cleaning rag. Wipe baseboards and window sills with dry cloth, then spray with cleaner and buff", "actions": ["[WALK] <bedroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <wall>", "[WALK] <window>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'couch', 'candle', 'clothes_hat', 'pillow', 'wall', 'oven', 'coffe_maker', 'television', 'curtain', 'diary', 'towel_rack', 'mouse', 'dishrack', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_pizza', 'dry_pasta', 'walllamp', 'trashcan', 'food_carrot', 'dough', 'knifeblock', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'toaster', 'wall', 'wall_clock', 'oven', 'food_cheese', 'food_food', 'coffe_maker', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher']\nTask:Keep an eye on stove as something is cooking\nresponse:\n", "output": {"thoughts": "Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <pot>", "[TURNTO] <pot>", "[LOOKAT] <pot>", "[FIND] <wall_clock>", "[TURNTO] <wall_clock>", "[LOOKAT] <wall_clock>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'cat', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'television', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'sink', 'bathroom_cabinet', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'dry_pasta', 'microwave', 'knifeblock', 'knife', 'longboard', 'ceiling', 'purse', 'face', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'towel', 'pillow', 'nightstand', 'powersocket', 'photoframe', 'bed', 'clothes_gloves', 'mouthwash', 'after_shave', 'shoes', 'dishrack', 'dresser', 'hair', 'home_office', 'table', 'dishwasher']\nTask:Rain welcome\nresponse:\n", "output": {"thoughts": "The robot would greet me and then hand me a towel.  It would make an appropriate quip about the weather while I dried off.  It would receive the wet towel when I was done.", "actions": ["[WALK] <bedroom>", "[WALK] <towel>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <face>", "[WIPE] <face>", "[FIND] <hair>", "[WIPE] <hair>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'band-aids', 'walllamp', 'drawing', 'wallshelf', 'floor', 'character', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'bench', 'photoframe', 'television', 'tray', 'curtain', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'home_office', 'table', 'clothes_shirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Sitting down and relaxing on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['scissors', 'tea', 'walllamp', 'knife', 'orchid', 'cpuscreen', 'hanger', 'character', 'desk', 'clothes_jacket', 'toaster', 'mat', 'tvstand', 'shower', 'clothes_hat', 'food_orange', 'bowl', 'bench', 'powersocket', 'oven', 'bed', 'light_bulb', 'coffe_maker', 'curtain', 'kitchen_counter', 'comb', 'pajamas', 'controller', 'sink', 'tooth_paste']\nTask:Grab things\nresponse:\n", "output": {"thoughts": "The robot will need some kind of hands or grabbing device to transport objects.", "actions": ["[WALK] <novel>", "[TURNTO] <novel>", "[GRAB] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'rag', 'doorjamb', 'cards', 'dresser', 'sink', 'dishwasher']\nTask:Watch  TV\nresponse:\n{\n    \"thoughts\": Sit on my couch  directly opposite  my tv , switch  on my tv with the  remote control  and watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'freezer', 'walllamp', 'ground_coffee', 'bathroom', 'ceilingfan', 'clothes_underwear', 'bathroom_counter', 'character', 'thread', 'piano_bench', 'toaster', 'deck_of_cards', 'homework', 'couch', 'dustpan', 'powersocket', 'newspaper', 'clothes_gloves', 'novel', 'coffe_maker', 'doorjamb', 'chessboard', 'cd', 'napkin', 'home_office', 'instrument_guitar', 'clothes_skirt', 'light']\nTask:Close door\nresponse:\n", "output": {"thoughts": "Push the door shut until it latches.", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'woman', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'man', 'bookshelf', 'keyboard', 'piano_bench', 'stovefan', 'tvstand', 'shower', 'faucet', 'oil', 'dining_room', 'food_kiwi', 'nightstand', 'bench', 'oven', 'powersocket', 'coffe_maker', 'tray', 'curtain', 'towel_rack', 'cupboard', 'door', 'sink', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Stand next to the door. When the door opens from the outside, greet the guests.,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <woman>', '[FIND] <woman>', '[TURNTO] <woman>', '[LOOKAT] <woman>', '[GREET] <woman>', '[FIND] <man>', '[TURNTO] <man>', '[LOOKAT] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['juice', 'measuring_cup', 'toilet', 'nail_polish', 'trashcan', 'dough', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'mop_bucket', 'bookmark', 'ceilinglamp', 'deck_of_cards', 'shoe-shine_kit', 'window', 'couch', 'wall', 'nightstand', 'check', 'mouthwash', 'television', 'hairbrush', 'sink', 'home_office', 'table', 'dishwasher', 'soap', 'teeth']\nTask:Do homework\nresponse:\n", "output": {"thoughts": "Sit on the couch in the Living room and pick u some papers", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <check>", "[GRAB] <check>", "[DROP] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'computer', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'television', 'curtain', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to Living room. Find Television. Walk to sofa. Sit on sofa. Find Remote. Grab remote. Turn to television. Look at television. Switch on Television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'walllamp', 'detergent', 'milk', 'hanger', 'phone', 'toaster', 'tvstand', 'balanceball', 'electric_shaver', 'food_noodles', 'garbage_can', 'mousepad', 'clothes_hat', 'pillow', 'broom', 'food_kiwi', 'bench', 'food_vegetable', 'food_cheese', 'food_food', 'ice', 'wine', 'rag', 'shaving_cream', 'dresser', 'cupboard', 'napkin', 'pantry']\nTask:Organize pantry\nresponse:\n", "output": {"thoughts": "Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.", "actions": ["[WALK] <dining_room>", "[WALK] <pantry>", "[FIND] <food_food>", "[GRAB] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[FIND] <garbage_can>", "[PUTBACK] <food_food> <garbage_can>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <dry_pasta>", "[GRAB] <dry_pasta>", "[TURNTO] <dry_pasta>", "[LOOKAT] <dry_pasta>", "[PUTOBJBACK] <dry_pasta>", "[FIND] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[PUSH] <food_food>", "[FIND] <food_noodles>", "[GRAB] <food_noodles>", "[TURNTO] <food_noodles>", "[LOOKAT] <food_noodles>", "[PUTOBJBACK] <food_noodles>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'orchid', 'bills', 'drawing', 'floor', 'ceiling', 'table_cloth', 'toaster', 'mat', 'faucet', 'mousepad', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'cup', 'keys', 'coffe_maker', 'food_food', 'check', 'water', 'wine', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'light', 'board_game']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Go to the kitchen and drink some water.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <cup> <sink>', '[FIND] <water>', '[GRAB] <cup>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>', '[PUTBACK] <cup> <sink>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'beer', 'chair', 'cpuscreen', 'bathroom_counter', 'phone', 'desk', 'cleaning_solution', 'bookmark', 'iron', 'mousepad', 'couch', 'dustpan', 'food_jam', 'tape', 'television', 'wooden_spoon', 'curtain', 'rag', 'doorjamb', 'hairbrush', 'cupboard', 'door', 'bathroom_cabinet', 'table', 'vase', 'colander', 'teeth']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray the mirror with a cleaner, take a dry rag and wipe the mirror down until its clean.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'iron', 'mousepad', 'oil', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'video_game_controller', 'television', 'doorjamb', 'kettle', 'mouse', 'door', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on Computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'detergent', 'milk', 'drawing', 'orchid', 'cpuscreen', 'longboard', 'floor', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'laptop', 'window', 'mousepad', 'clothes_socks', 'couch', 'wall', 'nightstand', 'powersocket', 'television', 'console', 'mouse', 'pajamas', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Use laptop\nresponse:\n", "output": {"thoughts": "Sit down on a couch and perform tasks such as navigating the internet on a laptop", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <keyboard>", "[LOOKAT] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'freezer', 'bathroom', 'detergent', 'food_butter', 'longboard', 'bookshelf', 'stovefan', 'shower', 'pillow', 'towel', 'dining_room', 'nightstand', 'powersocket', 'toy', 'basket_for_clothes', 'newspaper', 'light_bulb', 'photoframe', 'food_food', 'coin', 'doorjamb', 'headset', 'shoes', 'hairbrush', 'cupboard', 'napkin', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into my room and turn the radio on so that I can listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'knife', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'shoe-shine_kit', 'balanceball', 'homework', 'mousepad', 'wall', 'nightstand', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'light']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Keep my keyboard vacuumed and free of debris.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'microwave', 'bathroom', 'trashcan', 'toothbrush', 'drawing', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'purse', 'food_bacon', 'ceilinglamp', 'plate', 'laptop', 'deck_of_cards', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'spoon', 'home_office', 'table', 'teeth']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'beer', 'mop', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'balanceball', 'electric_shaver', 'mousepad', 'wall', 'nightstand', 'bed', 'doorjamb', 'towel_rack', 'mouse', 'hairdryer', 'maindoor', 'light']\nTask:Clean screen\nresponse:\n", "output": {"thoughts": "Use a soft rag to wipe the screen until there are no marks on it.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <computer>", "[WIPE] <computer>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'dough', 'drawing', 'needle', 'food_egg', 'wallshelf', 'cpuscreen', 'orchid', 'floor', 'ceiling', 'computer', 'mail', 'mat', 'tvstand', 'shoe-shine_kit', 'couch', 'pillow', 'dining_room', 'wall', 'photoframe', 'food_food', 'television', 'coin', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'clothes_dress', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, turn to the sofa. Sit on the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'walllamp', 'woman', 'drawing', 'hanger', 'longboard', 'bathroom_counter', 'phone', 'character', 'man', 'food_cake', 'alcohol', 'stovefan', 'window', 'mousepad', 'microphone', 'powersocket', 'cup', 'food_cheese', 'check', 'food_food', 'after_shave', 'controller', 'maindoor', 'food_steak', 'hairdryer', 'bathroom_cabinet', 'dirt']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Stand next to the door. When the door opens from the outside, greet the guests.", "actions": ["[RUN] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'pillow', 'wall', 'powersocket', 'photoframe', 'bed', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'sink', 'light', 'coffee']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room ,find television , switch on remote control,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[SWITCHON] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'food_pizza', 'beer', 'ground_coffee', 'bathroom', 'walllamp', 'ceilingfan', 'orchid', 'character', 'desk', 'cleaning_solution', 'bookshelf', 'food_bacon', 'mat', 'mousepad', 'wall', 'pasta', 'oven', 'powersocket', 'novel', 'food_rice', 'wooden_spoon', 'curtain', 'rag', 'console', 'chessboard', 'headset', 'dresser', 'cupboard']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Spray counter tops with all purpose cleaner. Using towel, wipe counter, removing all dirt and grime until clean.", "actions": ["[WALK] <dining_room>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'tea', 'band-aids', 'walllamp', 'nail_polish', 'knifeblock', 'bills', 'cpuscreen', 'hanger', 'shampoo', 'phone', 'floor', 'bedroom', 'desk', 'toaster', 'stereo', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bowl', 'dining_room', 'bench', 'newspaper', 'clothes_gloves', 'coffe_maker', 'curtain', 'diary', 'hairbrush', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['mop', 'walllamp', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'bookshelf', 'keyboard', 'mat', 'shower', 'faucet', 'couch', 'pillow', 'dining_room', 'photoframe', 'newspaper', 'novel', 'check', 'toilet_paper', 'coin', 'mirror', 'food_rice', 'rag', 'doorjamb', 'shoes', 'chessboard', 'dresser', 'hair', 'maindoor', 'table', 'clothes_shirt']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "Spray glass cleaner on mirror. Wipe with a clean cloth.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[WALK] <mirror>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'ceilingfan', 'food_carrot', 'drawing', 'floor', 'ceiling', 'phone', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'mat', 'oil', 'couch', 'dining_room', 'wall', 'bed', 'light_bulb', 'coffe_maker', 'television', 'curtain', 'kettle', 'door', 'napkin', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to my living room and switch on the television, and lay on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'bathroom', 'trashcan', 'ceilingfan', 'food_carrot', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'purse', 'computer', 'table_cloth', 'brush', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'doorjamb', 'mouse', 'napkin', 'light', 'vase']\nTask:Wash monitor\nresponse:\n", "output": {"thoughts": "Clean the grime off of my monitor screen when not in use", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>", "[FIND] <brush>", "[TURNTO] <brush>", "[LOOKAT] <brush>", "[GRAB] <brush>", "[WIPE] <computer>", "[DROP] <brush>", "[FIND] <napkin>", "[GRAB] <napkin>", "[WASH] <napkin>", "[WIPE] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'mat', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'food_food', 'television', 'doorjamb', 'mouse', 'cupboard', 'conditioner', 'door', 'bathroom_cabinet', 'light', 'table']\nTask:Look at painting\nresponse:\n{\n    \"thoughts\": Walk to the painting.  Look up at the painting.  Admire it on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <drawing>', '[FIND] <drawing>', '[TURNTO] <drawing>', '[LOOKAT] <drawing>']\n}\n", "input": "\nScene_Objects:['microwave', 'detergent', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'bookshelf', 'purse', 'food_cake', 'computer', 'ceilinglamp', 'mat', 'electric_shaver', 'pillow', 'dining_room', 'wall', 'toy', 'cup', 'television', 'vacuum_cleaner', 'shoes', 'comb', 'dresser', 'cupboard', 'food_steak', 'hairdryer', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Clean room\nresponse:\n", "output": {"thoughts": "Pick up his toys, vacuum the garbage off the floor.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTBACK] <toy> <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[PUTBACK] <toy> <dresser>", "[CLOSE] <dresser>", "[WALK] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ironing_board', 'sheets', 'tablelamp', 'detergent', 'knifeblock', 'bathroom_counter', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'iron', 'window', 'couch', 'pillow', 'bowl', 'nightstand', 'food_cheese', 'video_game_controller', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'pajamas', 'bag', 'dishwasher', 'board_game', 'clothes_shirt']\nTask:Iron shirt\nresponse:\n{\n    \"thoughts\": Set up an ironing board and an iron.  Lay a shirt on board.  Run iron over shirt until it is free of wrinkles.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <ironing_board>', '[FIND] <ironing_board>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <ironing_board>', '[FIND] <iron>', '[GRAB] <iron>', '[PLUGIN] <iron>', '[PLUGOUT] <iron>', '[PUTOBJBACK] <iron>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'crayon', 'toilet', 'chair', 'freezer', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'computer', 'clothes_jacket', 'ceilinglamp', 'window', 'pillow', 'wall', 'pasta', 'powersocket', 'toy', 'box', 'cup', 'check', 'coffe_maker', 'toothbrush_holder', 'doorjamb', 'dresser', 'door', 'light', 'scrabble']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTBACK] <toy> <dresser>", "[CLOSE] <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'crayon', 'microwave', 'dog', 'band-aids', 'walllamp', 'trashcan', 'knifeblock', 'orchid', 'floor', 'desk', 'bookshelf', 'computer', 'piano_bench', 'plate', 'faucet', 'dish_soap', 'food_jam', 'form', 'wall', 'photoframe', 'doorjamb', 'kettle', 'kitchen_counter', 'dishrack', 'cupboard', 'pajamas', 'sink', 'table']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walk to the kitchen and turn to the sink. Put the plate into the sink. Grab and squeeze the dish soap. Scrub the plate then rinse it. Put the plate into the dish rack.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <sink>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[SQUEEZE] <dish_soap>', '[SCRUB] <plate>', '[RINSE] <plate>', '[GRAB] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'band-aids', 'freezer', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'envelope', 'computer', 'bookmark', 'brush', 'ceilinglamp', 'mat', 'electric_shaver', 'mousepad', 'towel', 'wall', 'powersocket', 'light_bulb', 'novel', 'curtain', 'doorjamb', 'mouse', 'door', 'light', 'colander', 'board_game']\nTask:Read to child\nresponse:\n", "output": {"thoughts": "Read a book out loud to a child", "actions": ["[WALK] <bedroom>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[TURNTO] <novel>", "[POINTAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'mousepad', 'wall', 'nightstand', 'powersocket', 'coffe_maker', 'check', 'television', 'wine', 'doorjamb', 'mouse', 'cupboard', 'controller', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to my desk. I turn on the computer. I open a web page and look at it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['bathroom', 'detergent', 'cat', 'knife', 'food_egg', 'character', 'stamp', 'bedroom', 'desk', 'mat', 'ceilinglamp', 'shower', 'faucet', 'stereo', 'window', 'oil', 'mousepad', 'clothes_hat', 'towel', 'pillow', 'dining_room', 'powersocket', 'coin', 'console', 'towel_rack', 'dishrack', 'cupboard', 'maindoor', 'sink', 'bag']\nTask:Keep cats out of room\nresponse:\n", "output": {"thoughts": "Stand in the doorway and create a loud noise so the cats will stay out.", "actions": ["[WALK] <bedroom>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'beer', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'mat', 'stereo', 'couch', 'pillow', 'bowl', 'wall', 'nightstand', 'powersocket', 'cup', 'toilet_paper', 'television', 'after_shave', 'rag', 'doorjamb', 'coffee_pot', 'dishrack', 'instrument_guitar', 'table', 'soap', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'food_pizza', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'cleaning_solution', 'mail', 'brush', 'toaster', 'tvstand', 'balanceball', 'window', 'food_noodles', 'mousepad', 'food_peanut_butter', 'clothes_hat', 'dining_room', 'pasta', 'bench', 'photoframe', 'oven_mitts', 'ice', 'rag', 'dresser', 'cupboard', 'food_steak', 'maindoor', 'table']\nTask:Clean  mirror\nresponse:\n", "output": {"thoughts": "GET CLEANING CLOTH, SPRAY CLEANER ON CLOTH, CLEAN MIRROR", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[POUR] <cleaning_solution> <rag>", "[PUTOBJBACK] <cleaning_solution>", "[WALK] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[WIPE] <mirror>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'broom', 'wall', 'pasta', 'powersocket', 'shelf', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the computer and turn it on. I look at the computer screen. I move and click the mouse to select buttons on the screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[PUSH] <mouse>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'toilet', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'pillow', 'broom', 'wall', 'nightstand', 'television', 'ice', 'curtain', 'mouse', 'maindoor', 'table']\nTask:Straighten paintings on wall\nresponse:\n", "output": {"thoughts": "Go up to the painting and tilt it which ever way it is uneven until it looks straight.", "actions": ["[WALK] <home_office>", "[WALK] <wall>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[WALK] <drawing>", "[FIND] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TOUCH] <drawing>", "[PULL] <drawing>", "[PUSH] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>", "[TURNTO] <drawing>", "[LOOKAT] <drawing>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'chair', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'mat', 'ceilinglamp', 'tvstand', 'stereo', 'floor_lamp', 'mousepad', 'wall', 'nightstand', 'novel', 'video_game_controller', 'television', 'mouse', 'door', 'napkin']\nTask:Study\nresponse:\n{\n    \"thoughts\": Study norwegian,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['toilet', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'cleaning_solution', 'faucet', 'balanceball', 'window', 'clothes_socks', 'food_orange', 'wall', 'powersocket', 'bench', 'cutting_board', 'photoframe', 'novel', 'check', 'rag', 'lighter', 'mouse', 'kitchen_counter', 'hair', 'sink', 'bathroom_cabinet', 'home_office', 'tooth_paste', 'scrabble', 'dirt']\nTask:Wash sink\nresponse:\n", "output": {"thoughts": "Turn on faucet. Rub down sink.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[FIND] <sink>", "[POUR] <cleaning_solution> <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[SCRUB] <sink>", "[WASH] <sink>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'wallshelf', 'ceiling', 'phone', 'keyboard', 'thread', 'piano_bench', 'brush', 'mat', 'ceilinglamp', 'window', 'oil', 'closetdrawer', 'mousepad', 'food_peanut_butter', 'candle', 'nightstand', 'powersocket', 'newspaper', 'novel', 'television', 'doorjamb', 'kitchen_counter', 'comb', 'cupboard', 'home_office', 'table', 'instrument_violin']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Twist the lamp switch until the light goes on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'dog', 'blow_dryer', 'trashcan', 'orchid', 'bathroom_counter', 'longboard', 'ceiling', 'character', 'bedroom', 'computer', 'mop_bucket', 'stereo', 'shower', 'window', 'food_peanut_butter', 'food_orange', 'bench', 'oven', 'bed', 'toilet_paper', 'television', 'razor', 'curtain', 'towel_rack', 'cupboard', 'door', 'dishwasher', 'soap']\nTask:Put umbrella away\nresponse:\n", "output": {"thoughts": "Place the umbrella in the umbrella bucket point first.", "actions": ["[WALK] <home_office>", "[WALK] <mop_bucket>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'tablelamp', 'microwave', 'detergent', 'clothes_underwear', 'ceiling', 'character', 'desk', 'mail', 'table_cloth', 'bookmark', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'mousepad', 'candle', 'food_jam', 'food_orange', 'nightstand', 'powersocket', 'newspaper', 'coin', 'food_onion', 'pencil', 'dresser', 'home_office', 'dishwasher', 'tooth_paste', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['walllamp', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'bedroom', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'pillow', 'dining_room', 'wall', 'bench', 'toy', 'box', 'photoframe', 'clothes_gloves', 'novel', 'video_game_controller', 'cards', 'vacuum_cleaner', 'dresser', 'controller', 'maindoor', 'napkin', 'dishwasher', 'instrument_violin']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Bend down and pick up the toys off of the floor.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <dresser>", "[PUTIN] <toy> <dresser>", "[FIND] <toy>", "[WALK] <dresser>", "[GRAB] <toy>", "[PUTIN] <toy> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'scissors', 'crayon', 'toilet', 'jelly', 'walllamp', 'nail_polish', 'detergent', 'food_cake', 'alcohol', 'piano_bench', 'toaster', 'food_noodles', 'garbage_can', 'pillow', 'broom', 'oven', 'toy', 'food_vegetable', 'food_food', 'television', 'oven_mitts', 'doorjamb', 'pencil', 'headset', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'pantry']\nTask:Organize pantry\nresponse:\n{\n    \"thoughts\": Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <pantry>', '[FIND] <food_food>', '[GRAB] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[FIND] <garbage_can>', '[PUTBACK] <food_food> <garbage_can>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <dry_pasta>', '[GRAB] <dry_pasta>', '[TURNTO] <dry_pasta>', '[LOOKAT] <dry_pasta>', '[PUTOBJBACK] <dry_pasta>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[PUSH] <food_food>', '[FIND] <food_noodles>', '[GRAB] <food_noodles>', '[TURNTO] <food_noodles>', '[LOOKAT] <food_noodles>', '[PUTOBJBACK] <food_noodles>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'freezer', 'bathroom', 'detergent', 'ceilingfan', 'milk', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'food_rice', 'mouse', 'pencil', 'pajamas', 'maindoor', 'board_game', 'teeth', 'shoe_rack']\nTask:Wipe down baseboards please\nresponse:\n", "output": {"thoughts": "Use a rag to get all the dust off the baseboards please.", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <wall>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'freezer', 'ground_coffee', 'bathroom', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'bookshelf', 'button', 'mat', 'shower', 'mousepad', 'oil', 'closetdrawer', 'microphone', 'oven', 'cutting_board', 'check', 'centerpiece', 'diary', 'clothes_dress', 'clothes_scarf', 'napkin', 'tooth_paste']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the cat. Pet the cat with your hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'freezer', 'walllamp', 'bathroom', 'hanger', 'phone', 'desk', 'bookshelf', 'ceilinglamp', 'shoe-shine_kit', 'balanceball', 'faucet', 'couch', 'nightstand', 'wall', 'bench', 'powersocket', 'clothes_gloves', 'check', 'television', 'food_onion', 'after_shave', 'curtain', 'shaving_cream', 'mouse', 'maindoor', 'sink', 'bag', 'home_office']\nTask:Open front door\nresponse:\n", "output": {"thoughts": "Walk over to the front door, turn the knob, and open the door.", "actions": ["[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'tablelamp', 'tea', 'microwave', 'chair', 'drawing', 'orchid', 'floor', 'computer', 'food_bacon', 'table_cloth', 'stovefan', 'toaster', 'homework', 'mousepad', 'couch', 'powersocket', 'oven', 'bench', 'check', 'oven_mitts', 'coffee_filter', 'curtain', 'rag', 'doorjamb', 'kitchen_counter', 'shoes', 'bathtub', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'ground_coffee', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'shampoo', 'character', 'computer', 'ceilinglamp', 'mat', 'iron', 'plate', 'window', 'couch', 'dining_room', 'wall', 'powersocket', 'basket_for_clothes', 'photoframe', 'newspaper', 'video_game_controller', 'television', 'centerpiece', 'doorjamb', 'towel_rack', 'door', 'bathroom_cabinet', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on a couch, grab a book, open it, and scan the pages", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'scissors', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'knifeblock', 'orchid', 'hanger', 'floor', 'bedroom', 'tea_bag', 'envelope', 'clothes_jacket', 'toaster', 'balanceball', 'faucet', 'iron', 'bed', 'coffe_maker', 'wooden_spoon', 'cards', 'headset', 'dresser', 'napkin', 'spoon', 'vase', 'instrument_violin', 'laser_pointer', 'creditcard']\nTask:Receive credit card\nresponse:\n{\n    \"thoughts\": hear the dooble, go to door, take de envelope, open it with a scissor, take the creditcard,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <envelope>', '[FIND] <envelope>', '[GRAB] <envelope>', '[WALK] <home_office>', '[WALK] <scissors>', '[FIND] <scissors>', '[OPEN] <envelope>', '[FIND] <creditcard>', '[GRAB] <creditcard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'mop', 'chair', 'band-aids', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'laptop', 'toaster', 'foundation', 'pillow', 'wall', 'powersocket', 'bench', 'novel', 'doorjamb', 'mouse', 'kitchen_counter', 'door', 'napkin', 'light', 'soap', 'instrument_violin']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Grasp light switch, flip light switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'computer', 'button', 'brush', 'mat', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'wall_clock', 'standingmirror', 'coffe_maker', 'television', 'food_rice', 'wine', 'curtain', 'doorjamb', 'diary', 'pencil', 'door', 'bathroom_cabinet', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'detergent', 'trashcan', 'dough', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'thread', 'clothes_jacket', 'toaster', 'ceilinglamp', 'shoe-shine_kit', 'brush', 'wall', 'toilet_paper', 'television', 'wooden_spoon', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'cupboard', 'laundry_detergent', 'light', 'dishwasher', 'board_game', 'clothes_shirt', 'teeth']\nTask:Sweep hallway please\nresponse:\n", "output": {"thoughts": "Vacuum all the way up and down the hallway.  Get all the dirt and crumbs out of carpet please.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": I need to complete work on my computer, so I will be pulling out my chair, turning on my computer and begin my task.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['blender', 'walllamp', 'bathroom', 'trashcan', 'food_carrot', 'drawing', 'floor', 'phone', 'bedroom', 'stamp', 'keyboard', 'table_cloth', 'stovefan', 'ceilinglamp', 'window', 'couch', 'towel', 'wall', 'oven', 'bed', 'photoframe', 'clothes_gloves', 'television', 'doorjamb', 'laundry_detergent', 'door', 'maindoor', 'home_office', 'table', 'clothes_shirt']\nTask:Read magazine\nresponse:\n", "output": {"thoughts": "Please sit on the couch and read a magazine", "actions": ["[WALK] <home_office>", "[WALK] <newspaper>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <newspaper>", "[LOOKAT] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'newspaper', 'television', 'curtain', 'rag', 'doorjamb', 'mouse', 'door', 'bathtub', 'table']\nTask:Use computer\nresponse:\n{\n    \"thoughts\": Browse reddit,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'orchid', 'drawing', 'knifeblock', 'floor', 'folder', 'envelope', 'table_cloth', 'mop_bucket', 'stovefan', 'brush', 'ceilinglamp', 'window', 'foundation', 'couch', 'pillow', 'wall', 'powersocket', 'toy', 'newspaper', 'light_bulb', 'food_donut', 'wooden_spoon', 'doorjamb', 'door', 'maindoor', 'napkin', 'light']\nTask:Put toys away\nresponse:\n", "output": {"thoughts": "Pick up all the toys on the floor and put them in their correct storage bin or shelf.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'scissors', 'knife', 'hanger', 'bathroom_counter', 'floor', 'bedroom', 'desk', 'keyboard', 'piano_bench', 'stovefan', 'stereo', 'window', 'wall', 'powersocket', 'cutting_board', 'cup', 'bed', 'console', 'pencil', 'shoes', 'door', 'bathtub', 'bag', 'instrument_guitar', 'table', 'home_office', 'light', 'board_game']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into my room and turn the radio on so that I can listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'cat', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'food_bacon', 'stovefan', 'ceilinglamp', 'iron', 'dustpan', 'pillow', 'wall', 'toy', 'box', 'centerpiece', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cd', 'dresser', 'cupboard', 'light', 'table']\nTask:Vacuum\nresponse:\n", "output": {"thoughts": "Plug in the vacuum. Turn it on. Push it back and forth across the floor until all the debris is gone.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'shampoo', 'stovefan', 'mat', 'shower', 'window', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'newspaper', 'food_food', 'drinking_glass', 'water', 'curtain', 'doorjamb', 'dresser', 'food_steak', 'door', 'home_office', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'food_pizza', 'dog', 'freezer', 'bathroom', 'trashcan', 'milk', 'knife', 'orchid', 'floor', 'ceiling', 'bedroom', 'purse', 'toaster', 'ceilinglamp', 'faucet', 'balanceball', 'garbage_can', 'food_dessert', 'wall', 'nightstand', 'oven', 'tape', 'food_cheese', 'food_vegetable', 'novel', 'toothbrush_holder', 'ice', 'colander']\nTask:Rotate stock in refrigerator\nresponse:\n", "output": {"thoughts": "Throw out expired items, put older items in front.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[TURNTO] <food_bread>", "[LOOKAT] <food_bread>", "[PUTOBJBACK] <food_bread>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[FIND] <garbage_can>", "[PUTBACK] <food_cheese> <garbage_can>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[TURNTO] <food_cheese>", "[LOOKAT] <food_cheese>", "[PUTOBJBACK] <food_cheese>", "[FIND] <food_dessert>", "[TURNTO] <food_dessert>", "[LOOKAT] <food_dessert>", "[PUSH] <food_dessert>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[TURNTO] <food_vegetable>", "[LOOKAT] <food_vegetable>", "[PUTBACK] <food_vegetable> <garbage_can>", "[FIND] <milk>", "[GRAB] <milk>", "[TURNTO] <milk>", "[LOOKAT] <milk>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'shampoo', 'bedroom', 'bookshelf', 'thread', 'mop_bucket', 'tvstand', 'faucet', 'mousepad', 'towel', 'powersocket', 'bed', 'coffe_maker', 'television', 'oven_mitts', 'tray', 'mouse', 'kitchen_counter', 'dresser', 'door', 'bathtub', 'maindoor', 'home_office', 'bathroom_cabinet', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": Enter living room. Walk to telephone. Pick up from hook.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'filing_cabinet', 'beer', 'measuring_cup', 'chair', 'ground_coffee', 'knifeblock', 'longboard', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'mail', 'toaster', 'picture', 'dustpan', 'wall', 'bench', 'cutting_board', 'tape', 'cup', 'coffe_maker', 'coffee_filter', 'water', 'after_shave', 'dresser', 'maindoor', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[GRAB] <ground_coffee>", "[PUTOBJBACK] <ground_coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'chair', 'walllamp', 'bathroom', 'ground_coffee', 'ceiling', 'character', 'bedroom', 'desk', 'keyboard', 'bookshelf', 'toaster', 'tvstand', 'faucet', 'shower', 'foundation', 'mousepad', 'wall', 'cup', 'newspaper', 'bed', 'coffe_maker', 'doorjamb', 'dresser', 'conditioner', 'door', 'light', 'dishwasher', 'tooth_paste']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": take remote control, push \"change channel\" button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'food_ice_cream', 'beer', 'food_oatmeal', 'microwave', 'freezer', 'walllamp', 'ground_coffee', 'milk', 'knifeblock', 'drawing', 'food_egg', 'floor', 'ceiling', 'tea_bag', 'bookshelf', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'toaster', 'food_peanut_butter', 'candle', 'pillow', 'bowl', 'wall', 'bench', 'cup', 'newspaper', 'food_food', 'coffe_maker', 'check', 'food_onion', 'cards', 'kitchen_counter', 'cupboard', 'sink', 'table', 'dishwasher', 'laser_pointer']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Put oatmeal into a bowl.  Add milk.  Heat in a microwave.  Put pod of coffee into a coffee maker.  Activate.  Pour juice into a glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <bowl>", "[GRAB] <bowl>", "[FIND] <food_oatmeal>", "[GRAB] <food_oatmeal>", "[PUTBACK] <food_oatmeal> <bowl>", "[GRAB] <food_oatmeal>", "[PUTOBJBACK] <food_oatmeal>", "[CLOSE] <cupboard>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <bowl>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>", "[FIND] <microwave>", "[OPEN] <microwave>", "[PUTIN] <bowl> <microwave>", "[CLOSE] <microwave>", "[SWITCHON] <microwave>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTOBJBACK] <juice>", "[CLOSE] <freezer>", "[FIND] <table>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'toilet', 'freezer', 'bathroom', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'thread', 'stovefan', 'mat', 'faucet', 'towel', 'dining_room', 'wall', 'wall_clock', 'oven', 'bench', 'nightstand', 'check', 'coffee_filter', 'tray', 'shaving_cream', 'towel_rack', 'pencil', 'mouse', 'sink', 'bathroom_cabinet', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. Walk to the sink. Turn on tap. Get soap. Wash and rinse hands. Turn off tap. Dry hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[LOOKAT] <sink>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[RINSE] <soap>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'tablelamp', 'detergent', 'cat', 'knife', 'drawing', 'hanger', 'bathroom_counter', 'desk', 'food_cake', 'purse', 'envelope', 'piano_bench', 'stovefan', 'toaster', 'stereo', 'balanceball', 'floor_lamp', 'powersocket', 'love_seat', 'newspaper', 'bed', 'photoframe', 'video_game_controller', 'diary', 'maindoor', 'light', 'soap', 'board_game']\nTask:Read\nresponse:\n", "output": {"thoughts": "Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <newspaper>", "[GRAB] <newspaper>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[READ] <newspaper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'computer', 'food_bacon', 'stovefan', 'mat', 'tvstand', 'window', 'mousepad', 'form', 'dining_room', 'wall', 'cutting_board', 'bed', 'clothes_gloves', 'check', 'television', 'oven_mitts', 'curtain', 'mouse', 'hair', 'napkin', 'vase', 'instrument_violin', 'laser_pointer']\nTask:Gaze out window\nresponse:\n{\n    \"thoughts\": Look at the yard from the window,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <window>', '[FIND] <window>', '[TURNTO] <window>', '[LOOKAT] <window>']\n}\n", "input": "\nScene_Objects:['mop', 'bathroom', 'detergent', 'trashcan', 'food_carrot', 'knifeblock', 'floor', 'bedroom', 'thread', 'computer', 'brush', 'plate', 'picture', 'window', 'mousepad', 'clothes_hat', 'pillow', 'wall', 'box', 'bed', 'television', 'curtain', 'food_onion', 'kitchen_counter', 'maindoor', 'table', 'light', 'clothes_shirt', 'dirt', 'teeth']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Climb into bed, then read the novel", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'trashcan', 'drawing', 'orchid', 'ceiling', 'bedroom', 'mail', 'mat', 'tvstand', 'mousepad', 'dining_room', 'wall', 'wall_clock', 'basket_for_clothes', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'laundry_detergent', 'door', 'clothes_dress', 'bathroom_cabinet', 'table', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'bookmark', 'ceilinglamp', 'tvstand', 'balanceball', 'picture', 'mousepad', 'couch', 'wall', 'bed', 'food_food', 'tray', 'after_shave', 'rag', 'doorjamb', 'maindoor', 'bathroom_cabinet', 'dishwasher', 'vase']\nTask:Put away groceries\nresponse:\n", "output": {"thoughts": "Place groceries in refrigerator from shopping at the grocery store.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'fork', 'tablelamp', 'tea', 'band-aids', 'bathroom', 'drawing', 'hanger', 'floor', 'ceiling', 'dvd_player', 'desk', 'remote_control', 'computer', 'food_bacon', 'mat', 'ceilinglamp', 'window', 'mousepad', 'closetdrawer', 'couch', 'clothes_pants', 'pillow', 'wall', 'standingmirror', 'television', 'curtain', 'doorjamb', 'table', 'soap']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I push my body to be near to the table in front of me, without stand up. I move my right hand to near the TV remote control on the table. I take the TV remote control. I look the number on the remote control. I touch number 3 button. I put back the remote control to the table. I put my body back to the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[PLUGIN] <television>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[GRAB] <remote_control>', '[TOUCH] <remote_control>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'phone', 'desk', 'stamp', 'keyboard', 'bookshelf', 'computer', 'table_cloth', 'ceilinglamp', 'balanceball', 'faucet', 'mousepad', 'wall', 'nightstand', 'bench', 'check', 'doorjamb', 'mouse', 'controller', 'maindoor', 'light', 'dishwasher']\nTask:Research\nresponse:\n", "output": {"thoughts": "I read different articles online to gain a better understanding of a particular business, industry, method, terminology, etc.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <check>", "[GRAB] <check>", "[READ] <check>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'bathroom', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'shampoo', 'tvstand', 'faucet', 'stereo', 'clothes_socks', 'couch', 'clothes_pants', 'cutting_board', 'basket_for_clothes', 'photoframe', 'food_food', 'video_game_controller', 'coffe_maker', 'coffee_filter', 'television', 'tray', 'curtain', 'shaving_cream', 'laundry_detergent', 'maindoor', 'spoon', 'home_office', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'microwave', 'toilet', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'phone', 'food_cake', 'computer', 'ceilinglamp', 'shoe-shine_kit', 'iron', 'stereo', 'electric_shaver', 'homework', 'balanceball', 'picture', 'window', 'dining_room', 'powersocket', 'cup', 'photoframe', 'mirror', 'tray', 'pencil', 'shoes', 'conditioner', 'sink', 'creditcard']\nTask:Shave\nresponse:\n", "output": {"thoughts": "Pick up razor.  Push the button to turn it on.  Move it in a linear motion against the beard growth.  Make sure to get the bottom of your chin, your cheeks, and mustache area.", "actions": ["[WALK] <bathroom>", "[WALK] <electric_shaver>", "[FIND] <electric_shaver>", "[GRAB] <electric_shaver>", "[FIND] <mirror>", "[TURNTO] <mirror>", "[SWITCHON] <electric_shaver>", "[SWITCHOFF] <electric_shaver>", "[PUTOBJBACK] <electric_shaver>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'picture', 'homework', 'mousepad', 'wall', 'oven', 'toy', 'newspaper', 'check', 'television', 'mouse', 'laundry_detergent', 'conditioner', 'laser_pointer']\nTask:Push in desk chair\nresponse:\n{\n    \"thoughts\": Push the desk chair until it is under the desk.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[TOUCH] <chair>', '[PUSH] <chair>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'food_bread', 'scissors', 'microwave', 'freezer', 'milk', 'knifeblock', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'mail', 'food_bacon', 'ceilinglamp', 'shower', 'faucet', 'bowl', 'food_kiwi', 'wall', 'oven', 'cup', 'bed', 'coffe_maker', 'video_game_controller', 'diary', 'dishrack', 'controller', 'conditioner', 'home_office']\nTask:Get glass of milk\nresponse:\n", "output": {"thoughts": "Get an empty glass. Take milk from refrigerator and open it. Pour milk into glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[PUTOBJBACK] <milk>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'chair', 'freezer', 'detergent', 'floor', 'phone', 'ceiling', 'bedroom', 'keyboard', 'stovefan', 'shower', 'faucet', 'window', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'bench', 'photoframe', 'coffe_maker', 'check', 'television', 'food_food', 'ice', 'rag', 'towel_rack', 'kitchen_counter', 'cupboard']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'chair', 'walllamp', 'cat', 'trashcan', 'ceilingfan', 'drawing', 'bills', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'photoframe', 'fax_machine', 'doorjamb', 'mouse', 'light', 'dishwasher', 'printing_paper']\nTask:Print out papers\nresponse:\n", "output": {"thoughts": "Turn on printer and fill with blank paper. Use computer to click on printing option. Take printed page out of print.", "actions": ["[WALK] <home_office>", "[WALK] <fax_machine>", "[FIND] <fax_machine>", "[SWITCHON] <fax_machine>", "[FIND] <printing_paper>", "[GRAB] <printing_paper>", "[PUTBACK] <printing_paper> <fax_machine>", "[FIND] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>", "[GRAB] <printing_paper>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'doorjamb', 'towel_rack', 'pencil', 'shoes', 'dresser', 'sink', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go to the the living room, sit down on the sofa, and then I switch on the TV with the TV remote,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'toilet', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'shoe-shine_kit', 'couch', 'wall', 'video_game_controller', 'television', 'coin', 'curtain', 'kitchen_counter', 'headset', 'dresser', 'controller', 'table', 'dishwasher', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit down on sofa, get the remote, turn on tv, change channel to show I want to watch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'tvstand', 'foundation', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'coffe_maker', 'television', 'food_rice', 'doorjamb', 'shoes', 'dresser', 'door', 'light', 'clothes_skirt']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to room. I enter room. I find where the lightswitch is. I turn on the lights with the lightswitch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[TURNTO] <light>', '[LOOKAT] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['ironing_board', 'dry_pasta', 'food_salt', 'toilet', 'freezer', 'band-aids', 'nail_polish', 'detergent', 'food_carrot', 'knifeblock', 'longboard', 'clothes_jacket', 'mat', 'iron', 'mousepad', 'form', 'towel', 'dining_room', 'cup', 'bed', 'coffe_maker', 'video_game_controller', 'television', 'ice', 'doorjamb', 'towel_rack', 'home_office', 'light', 'vase', 'clothes_shirt']\nTask:Iron shirt\nresponse:\n", "output": {"thoughts": "Set up an ironing board and an iron.  Lay a shirt on board.  Run iron over shirt until it is free of wrinkles.", "actions": ["[WALK] <bedroom>", "[WALK] <ironing_board>", "[FIND] <ironing_board>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <ironing_board>", "[FIND] <iron>", "[GRAB] <iron>", "[PLUGIN] <iron>", "[PLUGOUT] <iron>", "[PUTOBJBACK] <iron>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'mop', 'walllamp', 'detergent', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'toaster', 'stereo', 'window', 'wall', 'bed', 'novel', 'coffe_maker', 'coffee_filter', 'water', 'kitchen_counter', 'shoes', 'dresser', 'cupboard', 'sink', 'bathtub', 'spoon', 'light', 'dishwasher', 'table', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Go to the kitchen and make coffee.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[PUTOBJBACK] <coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'freezer', 'detergent', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'couch', 'wall', 'bench', 'love_seat', 'food_food', 'check', 'television', 'coffe_maker', 'curtain', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit and watch the television", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <love_seat>", "[SIT] <love_seat>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'ground_coffee', 'detergent', 'trashcan', 'orchid', 'drawing', 'bedroom', 'tea_bag', 'bookshelf', 'envelope', 'alcohol', 'computer', 'ceilinglamp', 'mat', 'shower', 'deck_of_cards', 'food_noodles', 'microphone', 'bowl', 'wall', 'bench', 'powersocket', 'photoframe', 'television', 'dresser', 'napkin', 'table', 'clothes_shirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'food_salt', 'toilet', 'freezer', 'detergent', 'bills', 'orchid', 'hanger', 'ceiling', 'envelope', 'table_cloth', 'brush', 'mat', 'ceilinglamp', 'shower', 'stovefan', 'nightstand', 'powersocket', 'bench', 'photoframe', 'check', 'food_food', 'wooden_spoon', 'sink', 'napkin', 'instrument_guitar', 'table', 'coffee']\nTask:Pull up carpet\nresponse:\n", "output": {"thoughts": "Pry up a corner of your old dirty carpet.  Tug as hard as you can until it comes off the tacs. Pull up the pad. Try not to be sick at all the dirt.", "actions": ["[WALK] <home_office>", "[WALK] <mat>", "[FIND] <mat>", "[TOUCH] <mat>", "[PULL] <mat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'measuring_cup', 'walllamp', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'food_noodles', 'couch', 'broom', 'wall', 'newspaper', 'light_bulb', 'novel', 'television', 'curtain', 'doorjamb', 'lighter', 'cupboard', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['ceilingfan', 'toothbrush', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'longboard', 'mop_bucket', 'mat', 'tvstand', 'faucet', 'couch', 'clothes_pants', 'towel', 'wall', 'basket_for_clothes', 'photoframe', 'light_bulb', 'wooden_spoon', 'after_shave', 'coffee', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'laser_pointer', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Turn faucet clockwise to turn on water.  Place hands under stream of water, pick up soap, rub hands with soap.  Put soap back down.  Rub hands together with soap on them.  Rinse hands under water from faucet.  Turn faucet counterclockwise to turn water off.  Dry hands by rubbing them on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'blender', 'food_salt', 'chair', 'walllamp', 'nail_polish', 'bathroom', 'milk', 'drawing', 'wallshelf', 'clothes_underwear', 'bathroom_counter', 'floor', 'character', 'stamp', 'envelope', 'button', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'closetdrawer', 'form', 'coffe_maker', 'water', 'mouse', 'dresser', 'hairbrush', 'cupboard', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen, hold drinking glass, lift glass, pour water in drinking glass,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[FIND] <water>', '[GRAB] <water>', '[WALK] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['food_bread', 'scissors', 'crayon', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'bookmark', 'toaster', 'ceilinglamp', 'stovefan', 'faucet', 'window', 'homework', 'couch', 'wall', 'child', 'bench', 'photoframe', 'newspaper', 'coffe_maker', 'razor', 'television', 'kitchen_counter', 'cupboard', 'table', 'coffee']\nTask:Let baby learn how to walk\nresponse:\n", "output": {"thoughts": "The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TOUCH] <child>", "[GREET] <child>", "[FIND] <couch>", "[TURNTO] <couch>", "[POINTAT] <couch>", "[TURNTO] <child>", "[LOOKAT] <child>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'dog', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'form', 'wall', 'powersocket', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'mouse', 'cupboard', 'bathtub', 'sink', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": go to home office. turn the computer on. begin working,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'computer', 'ceilinglamp', 'faucet', 'teeth', 'towel', 'wall', 'nightstand', 'powersocket', 'basket_for_clothes', 'photoframe', 'light_bulb', 'toilet_paper', 'television', 'food_donut', 'tray', 'cards', 'towel_rack', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Hand washing\nresponse:\n", "output": {"thoughts": "Turn on the faucet. Soap up your hands. Then rinse under the running water of the faucet. Turn off faucet and dry hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'remote_control', 'thread', 'brush', 'ceilinglamp', 'tvstand', 'mat', 'toaster', 'clothes_hat', 'dining_room', 'wall', 'powersocket', 'cup', 'photoframe', 'television', 'doorjamb', 'pencil', 'dresser', 'door', 'home_office', 'table', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Change the channel on the TV in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'blow_dryer', 'cat', 'trashcan', 'toothbrush', 'drawing', 'floor', 'character', 'desk', 'wine_glass', 'ceilinglamp', 'plate', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'food_cheese', 'bed', 'food_food', 'check', 'television', 'coffee_filter', 'curtain', 'door', 'spoon', 'table']\nTask:Entertain\nresponse:\n", "output": {"thoughts": "Give guests wine, cheese, crackers and sit and talk", "actions": ["[WALK] <home_office>", "[WALK] <food_cheese>", "[FIND] <food_cheese>", "[FIND] <food_food>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <wine_glass>", "[GRAB] <wine_glass>", "[DRINK] <wine_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'toilet', 'bathroom', 'milk', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'towel', 'wall', 'bench', 'cup', 'photoframe', 'toothbrush_holder', 'water', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light', 'clothes_skirt', 'soap']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>', '[TURNTO] <shower>', '[FIND] <water>', '[RINSE] <water>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <water>', '[FIND] <towel>', '[GRAB] <towel>', '[SCRUB] <towel>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'sheets', 'food_ice_cream', 'food_salt', 'measuring_cup', 'microwave', 'toilet', 'detergent', 'ceilingfan', 'drawing', 'hanger', 'floor', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'plate', 'window', 'floor_lamp', 'mousepad', 'couch', 'pillow', 'wall', 'clothes_gloves', 'novel', 'check', 'television', 'food_rice', 'doorjamb', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'novel', 'television', 'fax_machine', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I walk to the home office. I sit on the chair. I switch on the computer. I look at my monitor. I open the browsing application by click the application using mouse. I type the book that I need to searching for (using keyboard). I found the book that I need. I read the book. I switch on the printer. I print the information that I get. I read carefully the book information that I got. I switch off the printer. I switch of the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <fax_machine>', '[SWITCHON] <fax_machine>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>', '[SWITCHOFF] <fax_machine>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['fork', 'crayon', 'chair', 'food_butter', 'food_carrot', 'ceilingfan', 'drawing', 'knife', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'thread', 'brush', 'plate', 'wall', 'bench', 'powersocket', 'oven', 'check', 'food_food', 'doorjamb', 'console', 'dresser', 'cupboard', 'door', 'maindoor', 'table', 'light']\nTask:Eat\nresponse:\n", "output": {"thoughts": "I get my prepared dinner set on the dining table, sit down on chair, pick up utensils, eat food", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <food_food> <table>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <knife>", "[GRAB] <knife>", "[FIND] <plate>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'water_glass', 'pot', 'spectacles', 'chair', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'ceiling', 'character', 'stovefan', 'mat', 'faucet', 'food_orange', 'wall', 'oven', 'cutting_board', 'food_food', 'water', 'tray', 'food_onion', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'instrument_guitar', 'home_office']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen, take a glass from the cupboard, fill with water and drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>', '[RINSE] <water_glass>', '[PUTBACK] <water_glass> <cupboard>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'walllamp', 'ground_coffee', 'trashcan', 'knifeblock', 'orchid', 'floor', 'ceiling', 'folder', 'plate', 'faucet', 'tvstand', 'mousepad', 'candle', 'wall', 'pasta', 'oven', 'colander', 'wine', 'tray', 'kitchen_counter', 'dresser', 'cupboard', 'hair', 'sink', 'vase', 'soap']\nTask:Wipe down counter\nresponse:\n", "output": {"thoughts": "Clean counter", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <rag>", "[SQUEEZE] <rag>", "[SWITCHOFF] <faucet>", "[WALK] <kitchen_counter>", "[WIPE] <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'woman', 'knifeblock', 'needle', 'cpuscreen', 'desk', 'man', 'keyboard', 'bedroom', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'nightstand', 'oven', 'cup', 'food_food', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['glue', 'food_bread', 'chair', 'freezer', 'band-aids', 'walllamp', 'food_butter', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'button', 'clothes_jacket', 'toaster', 'ceilinglamp', 'microphone', 'wall', 'oven', 'powersocket', 'bed', 'food_food', 'coffe_maker', 'oven_mitts', 'ice', 'tray', 'curtain', 'pencil', 'clothes_dress', 'dishwasher']\nTask:Make toast\nresponse:\n", "output": {"thoughts": "Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_bread>", "[GRAB] <food_bread>", "[CLOSE] <freezer>", "[FIND] <toaster>", "[PLUGIN] <toaster>", "[PUTBACK] <food_bread> <toaster>", "[SWITCHON] <toaster>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'novel', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'comb', 'door', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Find book, grab book, fin bookmark, sit down somewhere comfortable, open book, read book, close book after reading.,\n    \"actions\": ['[WALK] <desk>', '[FIND] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[FIND] <couch>', '[SIT] <couch>', '[GRAB] <novel>', '[READ] <novel>', '[PUTBACK] <bookmark> <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'crayon', 'measuring_cup', 'tea', 'mop', 'chair', 'nail_polish', 'bathroom', 'hanger', 'shampoo', 'phone', 'character', 'thread', 'tvstand', 'food_peanut_butter', 'bench', 'tape', 'food_food', 'mouthwash', 'razor', 'wooden_spoon', 'curtain', 'lighter', 'controller', 'laundry_detergent', 'door', 'bag', 'dishwasher', 'teeth']\nTask:Watch fly\nresponse:\n", "output": {"thoughts": "Look around for a fly, often on a wall or ceiling.  Follow it's movements.", "actions": ["[WALK] <bedroom>", "[WALK] <fly>", "[FIND] <fly>", "[TURNTO] <fly>", "[LOOKAT] <fly>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'crayon', 'tea', 'freezer', 'walllamp', 'bathroom', 'detergent', 'blow_dryer', 'drawing', 'floor', 'desk', 'thread', 'mop_bucket', 'tvstand', 'towel', 'wall', 'toy', 'cup', 'check', 'razor', 'food_donut', 'wine', 'centerpiece', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'laundry_detergent', 'light', 'soap']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Robot has a vacuum underneath its feat/body that will allow it to clean the surface of the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[WALK] <dining_room>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[WALK] <home_office>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <filing_cabinet>', '[PUTBACK] <vacuum_cleaner> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['trashcan', 'floor', 'ceiling', 'phone', 'computer', 'bookmark', 'toaster', 'mat', 'stereo', 'window', 'food_jam', 'pillow', 'dining_room', 'wall', 'powersocket', 'bed', 'check', 'mouthwash', 'television', 'tray', 'curtain', 'dishrack', 'dresser', 'laundry_detergent', 'hairdryer', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'shoe_rack']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I lie on my bed and read a book.", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <bed>", "[LIE] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'crayon', 'toilet', 'bathroom', 'nail_polish', 'trashcan', 'food_carrot', 'hanger', 'floor', 'computer', 'toaster', 'window', 'food_jam', 'nightstand', 'oven', 'bench', 'tape', 'photoframe', 'food_food', 'coffe_maker', 'television', 'toilet_paper', 'doorjamb', 'dresser', 'pajamas', 'cupboard', 'door', 'sink', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into the living room. I find the stereo and turn it on, and I listen to the music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['chair', 'bathroom', 'drawing', 'clothes_underwear', 'bathroom_counter', 'floor', 'character', 'computer', 'mop_bucket', 'bookmark', 'brush', 'ceilinglamp', 'faucet', 'window', 'couch', 'towel', 'dining_room', 'wall', 'newspaper', 'bed', 'television', 'wine', 'rag', 'mouse', 'napkin', 'bathroom_cabinet', 'table', 'dishwasher', 'light', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up a book and open to the last page you were on. Read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <bookmark>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'knifeblock', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'desk', 'food_cake', 'stovefan', 'toaster', 'deck_of_cards', 'faucet', 'candle', 'pillow', 'dining_room', 'oven', 'photoframe', 'check', 'food_food', 'television', 'bathroom_cabinet', 'table', 'dishwasher', 'soap', 'board_game', 'clothes_shirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find cat. Walk to cat. Pull cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[PULL] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'dry_pasta', 'food_salt', 'chair', 'drawing', 'bills', 'hanger', 'floor', 'phone', 'ceiling', 'toaster', 'plate', 'wall', 'bench', 'powersocket', 'tape', 'cup', 'bed', 'check', 'food_food', 'tray', 'doorjamb', 'pencil', 'dishrack', 'dresser', 'cupboard', 'door', 'table', 'light', 'laser_pointer']\nTask:Eat\nresponse:\n", "output": {"thoughts": "Serve food on dinning table in appropriate utensils, sit on the chair, eat", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[TURNTO] <table>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <plate>", "[PUTBACK] <food_food> <plate>", "[GRAB] <food_food>", "[PUTOBJBACK] <food_food>", "[FIND] <fork>", "[GRAB] <fork>", "[EAT] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'toilet', 'detergent', 'toothbrush', 'orchid', 'wallshelf', 'hanger', 'bedroom', 'bookshelf', 'computer', 'shower', 'window', 'cutting_board', 'food_cheese', 'novel', 'check', 'mouthwash', 'toothbrush_holder', 'wine', 'doorjamb', 'pajamas', 'hairdryer', 'maindoor', 'bathtub', 'home_office', 'door', 'clothes_skirt', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the toothbrush holder, grab the toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now wash the teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[TURNTO] <toothbrush_holder>', '[LOOKAT] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>']\n}\n", "input": "\nScene_Objects:['microwave', 'bathroom', 'cat', 'ceilingfan', 'woman', 'drawing', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'wall', 'child', 'oven', 'toy', 'bed', 'clothes_gloves', 'television', 'curtain', 'centerpiece', 'kitchen_counter', 'spoon', 'table', 'vase']\nTask:Watch movie\nresponse:\n", "output": {"thoughts": "Sitting and watching  a movie with family", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[TURNTO] <child>", "[LOOKAT] <child>", "[GREET] <child>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'scissors', 'tea', 'freezer', 'walllamp', 'knifeblock', 'knife', 'bills', 'floor', 'toaster', 'mat', 'plate', 'stereo', 'picture', 'tvstand', 'dish_soap', 'microphone', 'wall', 'oven', 'cup', 'light_bulb', 'coffe_maker', 'drinking_glass', 'tray', 'mouse', 'cupboard', 'controller', 'maindoor', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen and turn to the dishwasher, put the knife, glass, fork and plate into the dishwasher. Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[TURNTO] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[OPEN] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'trashcan', 'knifeblock', 'drawing', 'food_egg', 'clothes_underwear', 'orchid', 'bathroom_counter', 'ceiling', 'folder', 'cleaning_solution', 'mat', 'faucet', 'broom', 'wall', 'toy', 'box', 'photoframe', 'bed', 'food_food', 'mirror', 'paper_towel', 'dresser', 'maindoor', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'vase']\nTask:Clean mirror\nresponse:\n", "output": {"thoughts": "First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[CLOSE] <bathroom_cabinet>", "[WALK] <mirror>", "[FIND] <cleaning_solution>", "[RINSE] <cleaning_solution>", "[WIPE] <mirror>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mop_bucket', 'ceilinglamp', 'mat', 'electric_shaver', 'mousepad', 'wall', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'mouse', 'pencil', 'controller', 'door', 'table', 'teeth']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to home office ,find computer ,switch on computer and work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'video_game_console', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'dustpan', 'pillow', 'wall', 'cup', 'video_game_controller', 'television', 'curtain', 'dishrack', 'napkin', 'table', 'vase', 'scrabble', 'shoe_rack']\nTask:Play games\nresponse:\n", "output": {"thoughts": "Turn on game.  Use controller", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <video_game_console>", "[PLUGIN] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'cat', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'mousepad', 'bench', 'newspaper', 'coffe_maker', 'food_food', 'tray', 'curtain', 'mouse', 'door', 'hairdryer', 'sink', 'food_steak', 'light', 'dishwasher', 'spoon', 'table', 'board_game']\nTask:Write  school paper\nresponse:\n{\n    \"thoughts\": Writing  and drafting a paper for  class work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'food_salt', 'toilet', 'band-aids', 'drawing', 'orchid', 'clothes_underwear', 'hanger', 'longboard', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'ceilinglamp', 'mat', 'stereo', 'tvstand', 'towel', 'pillow', 'bowl', 'wall', 'nightstand', 'mouthwash', 'dresser', 'clothes_dress', 'maindoor', 'bathroom_cabinet', 'clothes_skirt', 'creditcard']\nTask:Hang up jacket\nresponse:\n", "output": {"thoughts": "I put clothes on hangers and hang them in the closet", "actions": ["[WALK] <bedroom>", "[WALK] <clothes_dress>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[WALK] <dresser>", "[FIND] <hanger>", "[OPEN] <dresser>", "[PUTIN] <clothes_dress> <dresser>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[WALK] <dresser>", "[PUTIN] <clothes_dress> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'spectacles', 'microwave', 'walllamp', 'drawing', 'orchid', 'hanger', 'floor', 'character', 'desk', 'alcohol', 'stovefan', 'ceilinglamp', 'mat', 'window', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'cutting_board', 'photoframe', 'light_bulb', 'television', 'curtain', 'kettle', 'napkin', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'toilet', 'bathroom', 'trashcan', 'milk', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'clothes_jacket', 'toaster', 'stovefan', 'tvstand', 'picture', 'window', 'dustpan', 'pillow', 'dining_room', 'wall', 'box', 'bed', 'wine', 'curtain', 'kettle', 'home_office', 'light', 'dishwasher']\nTask:Look out window\nresponse:\n", "output": {"thoughts": "I head into my child's room and look out the window to view the neighborhood from a different area in the house.", "actions": ["[WALK] <bedroom>", "[WALK] <window>", "[TURNTO] <window>", "[LOOKAT] <window>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'measuring_cup', 'toilet', 'chair', 'drawing', 'food_egg', 'clothes_underwear', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'bookshelf', 'button', 'shoe-shine_kit', 'shower', 'window', 'broom', 'wall', 'wall_clock', 'powersocket', 'bench', 'standingmirror', 'novel', 'coin', 'doorjamb', 'door', 'light', 'table', 'creditcard']\nTask:Read\nresponse:\n{\n    \"thoughts\": We pick out a book from the bookshelf and read it.,\n    \"actions\": ['[WALK] <novel>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'crayon', 'dog', 'toothbrush', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'faucet', 'shower', 'balanceball', 'mousepad', 'wall', 'pasta', 'powersocket', 'oven', 'nightstand', 'photoframe', 'video_game_controller', 'coffee_filter', 'wooden_spoon', 'console', 'shoes', 'sink', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "Put toothpaste on toothbrush, brush teeth, rinse mouth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>", "[PUTOBJBACK] <tooth_paste>", "[WALK] <sink>", "[FIND] <teeth>", "[SCRUB] <teeth>", "[SCRUB] <teeth>", "[SCRUB] <teeth>", "[PUTBACK] <toothbrush> <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[RINSE] <teeth>", "[SWITCHOFF] <faucet>", "[GRAB] <toothbrush>", "[PUTOBJBACK] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'walllamp', 'bathroom', 'cpuscreen', 'kitchen_cabinet', 'floor', 'tea_bag', 'toaster', 'mat', 'stereo', 'balanceball', 'dish_soap', 'towel', 'wall', 'oven', 'bed', 'light_bulb', 'novel', 'coffe_maker', 'check', 'coin', 'food_donut', 'tray', 'doorjamb', 'towel_rack', 'sink', 'dishwasher', 'vase', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTIN] <dish_soap> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'candle', 'wall', 'nightstand', 'powersocket', 'television', 'doorjamb', 'mouse', 'cupboard', 'sink', 'light', 'colander']\nTask:Work\nresponse:\n", "output": {"thoughts": "Sit at my computer desk, Turn on Computer. Get online. Work.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'phone', 'bookshelf', 'table_cloth', 'bookmark', 'toaster', 'mat', 'faucet', 'oven', 'cup', 'photoframe', 'check', 'television', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'door', 'sink', 'light', 'tooth_paste']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'scissors', 'microwave', 'ceilingfan', 'drawing', 'knifeblock', 'orchid', 'floor', 'ceiling', 'desk', 'bookshelf', 'purse', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'nightstand', 'child', 'cup', 'television', 'curtain', 'centerpiece', 'food_steak', 'hairdryer', 'table', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn on tv.  Sit with kids.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[FIND] <child>", "[GREET] <child>", "[FIND] <child>", "[GREET] <child>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>", "[TURNTO] <television>", "[WATCH] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'juice', 'microwave', 'toilet', 'chair', 'freezer', 'ground_coffee', 'bathroom', 'knife', 'ceiling', 'bedroom', 'clothes_jacket', 'bookmark', 'toaster', 'faucet', 'iron', 'couch', 'food_peanut_butter', 'dining_room', 'oven', 'cup', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find a cat, touch it gently,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'tablelamp', 'crayon', 'tea', 'mop', 'ground_coffee', 'detergent', 'drawing', 'floor', 'phone', 'ceilinglamp', 'laptop', 'shower', 'iron', 'window', 'oil', 'couch', 'pillow', 'wall', 'oven', 'cup', 'coffe_maker', 'toilet_paper', 'television', 'door', 'sink', 'table', 'scrabble', 'dirt']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Bring book, lie on the sofa, read book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[LIE] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'microwave', 'freezer', 'detergent', 'drawing', 'floor', 'ceiling', 'table_cloth', 'stovefan', 'ceilinglamp', 'iron', 'faucet', 'window', 'homework', 'bowl', 'wall', 'wall_clock', 'bench', 'shelf', 'newspaper', 'mouthwash', 'food_donut', 'doorjamb', 'coffee_pot', 'door', 'sink', 'bathroom_cabinet', 'light', 'board_game']\nTask:Turn light off\nresponse:\n{\n    \"thoughts\": Put the light switch into the down, or off, position.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'ceilinglamp', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'photoframe', 'mouthwash', 'television', 'doorjamb', 'mouse', 'chessboard', 'dresser', 'cupboard', 'door', 'light', 'table']\nTask:Re arrange office\nresponse:\n", "output": {"thoughts": "Walk in.Turn on light.Move furniture. Change the room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[PUSH] <desk>", "[WALK] <chair>", "[PUSH] <chair>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'trashcan', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'bookshelf', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'homework', 'pillow', 'wall', 'pasta', 'powersocket', 'photoframe', 'oven_mitts', 'doorjamb', 'kitchen_counter', 'comb', 'door', 'bathtub', 'light', 'dishwasher']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to bathroom, I open the door of the bathroom, and sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'toilet', 'walllamp', 'ceilingfan', 'knifeblock', 'drawing', 'floor', 'folder', 'bedroom', 'bookshelf', 'table_cloth', 'toaster', 'window', 'wall', 'pasta', 'oven', 'nightstand', 'cup', 'novel', 'coffe_maker', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'hair', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Take dishes out of dishwasher\nresponse:\n", "output": {"thoughts": "Take clean dishes out of the dishwasher and put away", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": I need to complete work on my computer, so I will be pulling out my chair, turning on my computer and begin my task.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'blender', 'food_salt', 'crayon', 'chair', 'band-aids', 'ceilingfan', 'needle', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'brush', 'ceilinglamp', 'shower', 'dining_room', 'wall', 'oven', 'powersocket', 'coffe_maker', 'curtain', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'maindoor', 'light']\nTask:Dust\nresponse:\n", "output": {"thoughts": "Using soft brush, carefully dust both Indonesian wedding doll statues.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <brush>", "[GRAB] <brush>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'desk', 'bedroom', 'stovefan', 'ceilinglamp', 'plate', 'mat', 'couch', 'wall', 'oven', 'food_food', 'tray', 'food_onion', 'curtain', 'diary', 'kitchen_counter', 'dishrack', 'sauce_pan', 'cupboard', 'spoon', 'home_office']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <spoon>', '[GRAB] <spoon>', '[FIND] <sauce_pan>', '[PUTBACK] <spoon> <sauce_pan>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <plate>', '[PUTBACK] <pot> <plate>']\n}\n", "input": "\nScene_Objects:['glue', 'microwave', 'dog', 'band-aids', 'jelly', 'walllamp', 'bathroom', 'ceilingfan', 'food_butter', 'bathroom_counter', 'video_game_console', 'tea_bag', 'keyboard', 'stovefan', 'mat', 'faucet', 'balanceball', 'couch', 'clothes_hat', 'form', 'nightstand', 'bench', 'oven', 'video_game_controller', 'tray', 'pencil', 'napkin', 'home_office', 'soap', 'dirt']\nTask:Playing video game\nresponse:\n", "output": {"thoughts": "Playing  video games  on various video game  systems such as wii u  and  xbox", "actions": ["[WALK] <home_office>", "[WALK] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>", "[PUTOBJBACK] <video_game_controller>", "[SWITCHOFF] <video_game_console>", "[FIND] <video_game_console>", "[SWITCHON] <video_game_console>", "[FIND] <video_game_controller>", "[GRAB] <video_game_controller>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'freezer', 'walllamp', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'toaster', 'shower', 'faucet', 'window', 'couch', 'wall', 'powersocket', 'bed', 'coffe_maker', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub', 'light', 'laser_pointer']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": walk to bathroom , find shower , take a shower,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'cloth_napkin', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'toilet_paper', 'coffee_filter', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'light', 'dishwasher', 'table']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Use a cloth to wipe away any marks on the computer screen.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <cloth_napkin>", "[GRAB] <cloth_napkin>", "[WIPE] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'knife', 'drawing', 'wallshelf', 'orchid', 'hanger', 'cpuscreen', 'floor', 'character', 'keyboard', 'purse', 'homework', 'mousepad', 'wall', 'pasta', 'bench', 'keys', 'television', 'wooden_spoon', 'towel_rack', 'kettle', 'kitchen_counter', 'hair', 'conditioner', 'home_office', 'table']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'tablelamp', 'blender', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'mat', 'balanceball', 'clothes_socks', 'dustpan', 'wall', 'powersocket', 'food_cheese', 'after_shave', 'doorjamb', 'towel_rack', 'dresser', 'door', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Hit light switch", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'scissors', 'freezer', 'band-aids', 'blow_dryer', 'bathroom', 'ceilingfan', 'drawing', 'bathroom_counter', 'ceiling', 'food_cake', 'toaster', 'mat', 'faucet', 'shower', 'mousepad', 'food_jam', 'clothes_pants', 'bowl', 'wall', 'photoframe', 'keys', 'novel', 'toilet_paper', 'tray', 'controller', 'sink', 'bathroom_cabinet', 'table']\nTask:Get toilet paper\nresponse:\n{\n    \"thoughts\": There are extra toilet papers in the bottom cabinet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <toilet_paper>', '[GRAB] <toilet_paper>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'chair', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'cup', 'bed', 'novel', 'television', 'curtain', 'after_shave', 'doorjamb', 'door', 'light']\nTask:Turn off light\nresponse:\n", "output": {"thoughts": "Hit the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toothbrush', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'mat', 'tvstand', 'towel', 'wall', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'coffee_filter', 'food_rice', 'after_shave', 'doorjamb', 'cards', 'door', 'napkin', 'light', 'vase']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk to the bedroom and use the button on the wall, next to the door to turn on the lights.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'scissors', 'food_ice_cream', 'microwave', 'freezer', 'ceilingfan', 'character', 'bookshelf', 'toaster', 'ceilinglamp', 'deck_of_cards', 'mat', 'tvstand', 'window', 'faucet', 'mousepad', 'oil', 'food_jam', 'wall', 'nightstand', 'newspaper', 'television', 'ice', 'centerpiece', 'cupboard', 'sink', 'napkin', 'home_office', 'dishwasher']\nTask:Get dressed\nresponse:\n", "output": {"thoughts": "I get dressed in my bedroom in the mornings.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PUTON] <shoes>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'needle', 'floor', 'phone', 'mail', 'mat', 'plate', 'faucet', 'stereo', 'couch', 'dustpan', 'dining_room', 'wall', 'bench', 'oven', 'toy', 'cup', 'newspaper', 'standingmirror', 'food_food', 'razor', 'console', 'towel_rack', 'bathtub', 'table']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'dog', 'freezer', 'walllamp', 'ground_coffee', 'knifeblock', 'floor', 'ceiling', 'tea_bag', 'envelope', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'couch', 'food_kiwi', 'wall', 'bed', 'newspaper', 'check', 'coffe_maker', 'food_food', 'curtain', 'kitchen_counter', 'cupboard', 'pajamas', 'food_snack', 'sink', 'coffee']\nTask:Have snack\nresponse:\n", "output": {"thoughts": "Pour soda into a glass. Get pretzels out of bag. Eat and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <juice>", "[GRAB] <juice>", "[POUR] <juice> <cup>", "[PUTIN] <juice> <freezer>", "[CLOSE] <freezer>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <food_snack>", "[GRAB] <food_snack>", "[CLOSE] <cupboard>", "[DRINK] <cup>", "[EAT] <food_snack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'walllamp', 'nail_polish', 'dough', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'cards', 'dishrack', 'door', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'blender', 'chair', 'food_carrot', 'drawing', 'hanger', 'floor', 'phone', 'keyboard', 'purse', 'button', 'clothes_jacket', 'toaster', 'faucet', 'candle', 'form', 'broom', 'wall', 'bench', 'powersocket', 'doorjamb', 'kettle', 'kitchen_counter', 'shoes', 'dresser', 'cupboard', 'door', 'hairdryer', 'table', 'light']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Walk in.Wipe off table.Pull out chairs.Wipe out chairs.Sweep floor.Put back chairs.Clean windows.Polish silverware.", "actions": ["[WALK] <dining_room>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <table>", "[FIND] <chair>", "[PULL] <chair>", "[WIPE] <chair>", "[PUSH] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[WIPE] <chair>", "[PUSH] <chair>", "[WIPE] <table>", "[PUTOBJBACK] <rag>", "[FIND] <broom>", "[GRAB] <broom>", "[PUTOBJBACK] <broom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'food_carrot', 'drawing', 'hanger', 'phone', 'ceiling', 'desk', 'computer', 'toaster', 'mat', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'towel', 'cup', 'food_cheese', 'keys', 'tray', 'after_shave', 'towel_rack', 'mouse', 'cupboard', 'sink', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room and I pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['juice', 'bathroom', 'detergent', 'nail_polish', 'trashcan', 'food_butter', 'orchid', 'hanger', 'longboard', 'floor', 'bathroom_counter', 'remote_control', 'envelope', 'computer', 'window', 'electric_shaver', 'wall', 'cutting_board', 'bed', 'television', 'tray', 'curtain', 'after_shave', 'lighter', 'mouse', 'cupboard', 'clothes_dress', 'light', 'clothes_skirt', 'board_game']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watching tv with a remote is always fun; lay back, relax and enjoy;", "actions": ["[WALK] <bedroom>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'computer', 'tvstand', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'cup', 'television', 'doorjamb', 'mouse', 'shoes', 'door', 'sink', 'napkin', 'clothes_scarf', 'table', 'light', 'dishwasher', 'colander']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room ,find sofa ,sit on sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['dog', 'freezer', 'bathroom', 'detergent', 'shampoo', 'desk', 'stamp', 'food_bacon', 'brush', 'ceilinglamp', 'oil', 'soap', 'food_orange', 'towel', 'nightstand', 'video_game_controller', 'razor', 'television', 'wine', 'food_onion', 'mouse', 'shoes', 'hair', 'door', 'sink', 'home_office', 'light', 'vase', 'colander', 'teeth']\nTask:Open door\nresponse:\n", "output": {"thoughts": "I walk up to the door.  I use my hand to turn the door knob to the left.  I pull the door open with the same hand I used to turn the door knob", "actions": ["[RUN] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'toilet', 'walllamp', 'ceilingfan', 'knifeblock', 'drawing', 'floor', 'folder', 'bedroom', 'bookshelf', 'table_cloth', 'toaster', 'window', 'wall', 'pasta', 'oven', 'nightstand', 'cup', 'novel', 'coffe_maker', 'tray', 'curtain', 'towel_rack', 'kitchen_counter', 'hair', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Take dishes out of dishwasher\nresponse:\n{\n    \"thoughts\": Take clean dishes out of the dishwasher and put away,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[CLOSE] <dishwasher>']\n}\n", "input": "\nScene_Objects:['glue', 'trashcan', 'orchid', 'knifeblock', 'floor', 'tea_bag', 'keyboard', 'computer', 'stovefan', 'tvstand', 'stereo', 'window', 'dustpan', 'candle', 'food_jam', 'wall', 'pasta', 'bed', 'photoframe', 'coffe_maker', 'television', 'tray', 'curtain', 'wooden_spoon', 'towel_rack', 'pencil', 'chessboard', 'cupboard', 'maindoor', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Put iPhone into speaker holder, open up the Pandora app, listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <stereo>", "[PLUGIN] <phone>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sheets', 'drawing', 'cpuscreen', 'ceiling', 'desk', 'bookmark', 'mat', 'plate', 'stereo', 'deck_of_cards', 'mousepad', 'couch', 'clothes_hat', 'clothes_pants', 'bench', 'clothes_gloves', 'video_game_controller', 'curtain', 'towel_rack', 'cd', 'dishrack', 'chessboard', 'pajamas', 'laundry_detergent', 'maindoor', 'door', 'sink', 'bathroom_cabinet', 'clothes_shirt']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find the radio; switch on the radio and listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'pillow', 'wall', 'nightstand', 'photoframe', 'wine', 'doorjamb', 'mouse', 'dresser', 'napkin', 'bathroom_cabinet', 'light']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Wipe down computer desk and monitor", "actions": ["[WALK] <home_office>", "[WALK] <rag>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <desk>", "[FIND] <computer>", "[WIPE] <computer>", "[WIPE] <desk>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'dry_pasta', 'ceilingfan', 'trashcan', 'drawing', 'longboard', 'floor', 'ceiling', 'bedroom', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'shower', 'deck_of_cards', 'couch', 'food_kiwi', 'wall', 'tape', 'basket_for_clothes', 'keys', 'coffe_maker', 'toilet_paper', 'television', 'curtain', 'cd', 'controller', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch. Get comfortable. Switch on the television. Stare at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'chair', 'drawing', 'needle', 'hanger', 'ceiling', 'phone', 'toaster', 'ceilinglamp', 'laptop', 'iron', 'mat', 'deck_of_cards', 'mousepad', 'oil', 'wall', 'basket_for_clothes', 'food_donut', 'food_rice', 'tray', 'curtain', 'doorjamb', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'light']\nTask:Put on coat\nresponse:\n", "output": {"thoughts": "I remove my scarf from a coat peg in the entrance hall.  I wrap the scarf around my neck", "actions": ["[WALK] <home_office>", "[WALK] <hanger>", "[TURNTO] <hanger>", "[LOOKAT] <hanger>", "[FIND] <clothes_scarf>", "[GRAB] <clothes_scarf>", "[PUTON] <clothes_scarf>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'mop', 'chair', 'detergent', 'cat', 'phone', 'stamp', 'keyboard', 'stovefan', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'deck_of_cards', 'toaster', 'mousepad', 'closetdrawer', 'dining_room', 'food_kiwi', 'wall', 'check', 'coffe_maker', 'curtain', 'rag', 'mouse', 'comb', 'bathroom_cabinet', 'dishwasher', 'soap']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I go looking for the cat in the living room where he usually sleeps. Once I find the cat I just pet him for a few minutes.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'tea', 'chair', 'freezer', 'ground_coffee', 'drawing', 'orchid', 'floor', 'purse', 'faucet', 'balanceball', 'food_noodles', 'food_peanut_butter', 'wall', 'powersocket', 'toy', 'bed', 'food_food', 'check', 'food_donut', 'doorjamb', 'pencil', 'laundry_detergent', 'door', 'bag', 'napkin', 'light', 'table']\nTask:Pick up toys\nresponse:\n", "output": {"thoughts": "Grab toys off floor.  Put it in a container.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'toothbrush', 'drawing', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bedroom', 'mat', 'faucet', 'shower', 'food_noodles', 'form', 'wall', 'wall_clock', 'oven', 'tape', 'photoframe', 'keys', 'toothbrush_holder', 'rag', 'doorjamb', 'sink', 'bathroom_cabinet', 'instrument_guitar', 'dishwasher', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I walk in to the bathroom. I turn on the faucet. I find the toothbrush and toothpaste. I grab the toothbrush and toothpaste. I rinse off the toothbrush. I add toothpaste to the toothbrush. I put back the toothpaste. I wash my teeth with the toothbrush. I rinse off the toothbrush. I put the toothbrush back. I rinse my mouth out. I turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[RINSE] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>', '[PUTOBJBACK] <tooth_paste>', '[FIND] <teeth>', '[WASH] <teeth>', '[RINSE] <toothbrush>', '[PUTOBJBACK] <toothbrush>', '[RINSE] <teeth>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['glue', 'sponge', 'scissors', 'chair', 'dog', 'drawing', 'bills', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'ceilinglamp', 'shoe-shine_kit', 'window', 'foundation', 'couch', 'wall', 'bench', 'basket_for_clothes', 'bed', 'novel', 'television', 'kitchen_counter', 'dresser', 'cupboard', 'bathroom_cabinet', 'light', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Grab book.  Read it.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <novel>", "[GRAB] <novel>", "[CLOSE] <dresser>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'cat', 'trashcan', 'milk', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'dvd_player', 'character', 'desk', 'bookshelf', 'food_cake', 'computer', 'brush', 'ceilinglamp', 'plate', 'deck_of_cards', 'wall', 'powersocket', 'toy', 'newspaper', 'bed', 'coffe_maker', 'curtain', 'pencil', 'hair', 'door', 'light']\nTask:Give milk to cat\nresponse:\n{\n    \"thoughts\": Grab milk and pour it into the plate so that the cat can drink,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[POINTAT] <cat>', '[PULL] <cat>', '[FIND] <milk>', '[GRAB] <milk>', '[FIND] <plate>', '[POUR] <milk> <plate>', '[TOUCH] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'food_ice_cream', 'crayon', 'cat', 'woman', 'knife', 'longboard', 'ceiling', 'tea_bag', 'keyboard', 'bookshelf', 'envelope', 'clothes_jacket', 'bookmark', 'ceilinglamp', 'homework', 'foundation', 'food_jam', 'coffe_maker', 'check', 'mouse', 'controller', 'laundry_detergent', 'maindoor', 'clothes_scarf', 'bathroom_cabinet', 'vase', 'clothes_shirt', 'scrabble']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet people into my home", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'measuring_cup', 'dog', 'walllamp', 'bathroom', 'floor', 'phone', 'bedroom', 'mat', 'stereo', 'faucet', 'window', 'closetdrawer', 'dining_room', 'food_cheese', 'standingmirror', 'bed', 'check', 'television', 'coffee_filter', 'food_rice', 'ice', 'curtain', 'doorjamb', 'pencil', 'conditioner', 'bathtub', 'napkin', 'vase']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'detergent', 'ceilingfan', 'trashcan', 'drawing', 'knife', 'floor', 'ceiling', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'picture', 'couch', 'wall', 'pasta', 'novel', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'maindoor', 'table', 'light', 'vase', 'tooth_paste']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I sit on my couch and watch TV in my living room", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TOUCH] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[WATCH] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'detergent', 'food_butter', 'dough', 'food_carrot', 'drawing', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'keyboard', 'computer', 'clothes_jacket', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'mousepad', 'towel', 'wall', 'oven', 'cup', 'photoframe', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap and hot water,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'microwave', 'freezer', 'detergent', 'drawing', 'floor', 'ceiling', 'stamp', 'bedroom', 'thread', 'computer', 'toaster', 'faucet', 'mousepad', 'food_peanut_butter', 'wall', 'powersocket', 'toy', 'cup', 'newspaper', 'doorjamb', 'chessboard', 'comb', 'door', 'sink', 'bag', 'light', 'board_game']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Enter inside kid's room, begin picking up toys from the floor and putting them in a toy box, leave once once.", "actions": ["[WALK] <bedroom>", "[WALK] <toy>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[FIND] <toy>", "[GRAB] <toy>", "[WALK] <filing_cabinet>", "[PUTBACK] <toy> <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'freezer', 'bathroom', 'cpuscreen', 'desk', 'tea_bag', 'alcohol', 'tvstand', 'oven', 'tape', 'cutting_board', 'cup', 'clothes_gloves', 'novel', 'toilet_paper', 'television', 'coffee_filter', 'wooden_spoon', 'mouse', 'cd', 'controller', 'maindoor', 'sink', 'home_office', 'light', 'dishwasher', 'soap', 'dirt', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room. Turn on radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'tablelamp', 'food_salt', 'dog', 'walllamp', 'trashcan', 'dough', 'bills', 'knifeblock', 'floor', 'computer', 'bookmark', 'window', 'couch', 'food_jam', 'pillow', 'wall', 'nightstand', 'oven', 'cutting_board', 'bed', 'photoframe', 'novel', 'food_food', 'television', 'curtain', 'diary', 'mouse', 'table']\nTask:Breakfast\nresponse:\n", "output": {"thoughts": "I have breakfast in bed, watching tv.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[GRAB] <food_food>", "[WALK] <bedroom>", "[WALK] <television>", "[SWITCHON] <television>", "[WALK] <bed>", "[LIE] <bed>", "[TURNTO] <television>", "[WATCH] <television>", "[EAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'food_salt', 'walllamp', 'bathroom', 'trashcan', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'tvstand', 'shower', 'faucet', 'towel', 'wall', 'cup', 'keys', 'food_food', 'television', 'water', 'curtain', 'lighter', 'kitchen_counter', 'shoes', 'door', 'table', 'light', 'soap', 'teeth']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": go to a bathroom, open the shower, let water wash your body,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <soap>', '[SCRUB] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'microwave', 'toilet', 'jelly', 'walllamp', 'trashcan', 'knifeblock', 'knife', 'desk', 'keyboard', 'computer', 'button', 'brush', 'window', 'mousepad', 'couch', 'bench', 'cutting_board', 'novel', 'check', 'coffe_maker', 'television', 'toothbrush_holder', 'food_rice', 'curtain', 'headset', 'dresser', 'home_office', 'table', 'dishwasher']\nTask:Lock door\nresponse:\n", "output": {"thoughts": "Turn the lock.", "actions": ["[WALK] <bedroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'juice', 'cd_player', 'chair', 'freezer', 'detergent', 'trashcan', 'knife', 'knifeblock', 'drawing', 'floor', 'ceiling', 'bedroom', 'keyboard', 'remote_control', 'computer', 'button', 'ceilinglamp', 'faucet', 'mousepad', 'wall', 'nightstand', 'after_shave', 'door', 'bathtub', 'home_office', 'table', 'clothes_skirt', 'instrument_guitar', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.,\n    \"actions\": ['[WALK] <cd_player>', '[SWITCHON] <cd_player>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'mop', 'toilet', 'chair', 'trashcan', 'knifeblock', 'hanger', 'longboard', 'floor', 'phone', 'keyboard', 'balanceball', 'foundation', 'pillow', 'wall', 'powersocket', 'box', 'photoframe', 'bed', 'check', 'rag', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'maindoor', 'light', 'colander']\nTask:Sit\nresponse:\n", "output": {"thoughts": "We sit at the dining room table", "actions": ["[WALK] <dining_room>", "[WALK] <chair>", "[FIND] <chair>", "[PULL] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'drawing', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'food_food', 'coffe_maker', 'television', 'tray', 'curtain', 'mouse', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'sink', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I sit in the sofa and I relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'powersocket', 'oven', 'coffe_maker', 'television', 'coffee_filter', 'food_rice', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'dirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I like to set on my sofa.Turn on the tv. And watch Perry Mason. Monday's through Friday's.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'television', 'coin', 'doorjamb', 'mouse', 'chessboard', 'bathtub', 'table']\nTask:Write  school paper\nresponse:\n{\n    \"thoughts\": Writing  and drafting a paper for  class work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'curtain', 'shoes', 'cd', 'vacuum_cleaner', 'dresser', 'cupboard', 'clothes_dress', 'home_office', 'table', 'shoe_rack']\nTask:Take off shoes\nresponse:\n", "output": {"thoughts": "Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.", "actions": ["[WALK] <bedroom>", "[WALK] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[OPEN] <dresser>", "[PUTIN] <shoes> <dresser>", "[FIND] <shoes>", "[GRAB] <shoes>", "[PULL] <shoes>", "[WALK] <dresser>", "[PUTIN] <shoes> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'bathroom', 'milk', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'envelope', 'computer', 'mat', 'ceilinglamp', 'stereo', 'mousepad', 'wall', 'toothbrush_holder', 'food_rice', 'ice', 'after_shave', 'mouse', 'headset', 'dresser', 'door', 'home_office', 'table', 'board_game']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office, walk to desk, sit in chair, turn on computer, look at computer, click mouse and type on keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'dry_pasta', 'ceilingfan', 'trashcan', 'drawing', 'longboard', 'floor', 'ceiling', 'bedroom', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'shower', 'deck_of_cards', 'couch', 'food_kiwi', 'wall', 'tape', 'basket_for_clothes', 'keys', 'coffe_maker', 'toilet_paper', 'television', 'curtain', 'cd', 'controller', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch. Get comfortable. Switch on the television. Stare at the television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'chair', 'cat', 'hanger', 'bathroom_counter', 'floor', 'bookshelf', 'keyboard', 'ceilinglamp', 'shower', 'foundation', 'couch', 'food_orange', 'powersocket', 'bench', 'clothes_gloves', 'toilet_paper', 'razor', 'food_donut', 'mouse', 'kitchen_counter', 'shoes', 'dishrack', 'dresser', 'pajamas', 'laundry_detergent', 'clothes_scarf', 'home_office', 'laser_pointer', 'creditcard']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Pour water into a glass and drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'freezer', 'ground_coffee', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'bedroom', 'stamp', 'computer', 'ceilinglamp', 'balanceball', 'oil', 'couch', 'broom', 'wall', 'oven', 'box', 'food_cheese', 'television', 'doorjamb', 'towel_rack', 'kitchen_counter', 'sink', 'vase', 'shoe_rack']\nTask:Eat cheese\nresponse:\n", "output": {"thoughts": "Open fridge.  Get cheese.  Eat it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_cheese>", "[GRAB] <food_cheese>", "[EAT] <food_cheese>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'microwave', 'toilet', 'walllamp', 'drawing', 'needle', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'desk', 'computer', 'bookmark', 'stovefan', 'plate', 'tvstand', 'dish_soap', 'closetdrawer', 'dining_room', 'wall', 'wall_clock', 'oven', 'powersocket', 'curtain', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'chair', 'jelly', 'food_carrot', 'cpuscreen', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'laptop', 'plate', 'mousepad', 'dining_room', 'wall', 'wine', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk in. Turn on the light. Set in my chair. Turn on lap top. And start to work.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <laptop>", "[GRAB] <laptop>", "[SWITCHON] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'trashcan', 'toothbrush', 'drawing', 'knifeblock', 'wallshelf', 'bathroom_counter', 'ceiling', 'phone', 'bedroom', 'stovefan', 'tvstand', 'food_peanut_butter', 'nightstand', 'wall', 'oven', 'powersocket', 'coffe_maker', 'toothbrush_holder', 'after_shave', 'curtain', 'coffee_pot', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'bathtub', 'sink', 'home_office', 'dishwasher', 'tooth_paste']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <tooth_paste> <toothbrush_holder>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <toothbrush>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'food_ice_cream', 'walllamp', 'bathroom', 'ceilingfan', 'trashcan', 'woman', 'drawing', 'orchid', 'food_egg', 'bathroom_counter', 'floor', 'bedroom', 'man', 'ceilinglamp', 'window', 'couch', 'wall', 'bench', 'cup', 'newspaper', 'novel', 'check', 'television', 'towel_rack', 'clothes_dress', 'table', 'clothes_skirt', 'creditcard']\nTask:Hang with friends\nresponse:\n", "output": {"thoughts": "Sit with friends. Talk", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[FIND] <woman>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <man>", "[LOOKAT] <man>", "[GREET] <man>", "[TURNTO] <woman>", "[LOOKAT] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'food_salt', 'toilet', 'walllamp', 'bathroom', 'food_butter', 'drawing', 'hanger', 'shampoo', 'bedroom', 'purse', 'computer', 'stovefan', 'laptop', 'ceilinglamp', 'electric_shaver', 'window', 'homework', 'clothes_socks', 'wall_clock', 'shelf', 'basket_for_clothes', 'bed', 'coffe_maker', 'check', 'wooden_spoon', 'centerpiece', 'lighter', 'kettle']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'beer', 'freezer', 'ground_coffee', 'nail_polish', 'bills', 'drawing', 'knifeblock', 'floor', 'folder', 'bedroom', 'tea_bag', 'mail', 'computer', 'towel', 'wall', 'powersocket', 'cutting_board', 'cup', 'keys', 'clothes_gloves', 'check', 'razor', 'tray', 'doorjamb', 'pencil', 'dresser', 'door', 'light', 'table']\nTask:Put clothes away\nresponse:\n", "output": {"thoughts": "I take clean clothes and either place them in dresser drawers or hang them in a closet.", "actions": ["[WALK] <bedroom>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[CLOSE] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'freezer', 'walllamp', 'cpuscreen', 'wallshelf', 'ceiling', 'computer', 'shower', 'window', 'mousepad', 'couch', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'bench', 'cup', 'bed', 'photoframe', 'check', 'coffe_maker', 'razor', 'mouse', 'cupboard', 'laundry_detergent', 'door', 'home_office', 'light', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'microwave', 'chair', 'band-aids', 'woman', 'needle', 'longboard', 'man', 'keyboard', 'envelope', 'toaster', 'mat', 'ceilinglamp', 'shower', 'faucet', 'stereo', 'nightstand', 'box', 'light_bulb', 'coffe_maker', 'curtain', 'towel_rack', 'shoes', 'dresser', 'hairbrush', 'door', 'vase', 'instrument_violin', 'creditcard']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.", "actions": ["[WALK] <home_office>", "[WALK] <man>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'ground_coffee', 'walllamp', 'food_butter', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'thread', 'stovefan', 'toaster', 'tvstand', 'faucet', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'coffe_maker', 'food_food', 'oven_mitts', 'coffee_filter', 'water', 'wine', 'wooden_spoon', 'controller', 'bathroom_cabinet', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['fork', 'toilet', 'trashcan', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'bookshelf', 'thread', 'computer', 'piano_bench', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'oven', 'bed', 'novel', 'check', 'television', 'curtain', 'towel_rack', 'dresser', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Put child in bed, sit down,read book.", "actions": ["[WALK] <bedroom>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <bed>", "[SIT] <bed>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'filing_cabinet', 'scissors', 'chair', 'ceilingfan', 'dough', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'coffee_table', 'television', 'coin', 'wooden_spoon', 'curtain', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'fork', 'filing_cabinet', 'toilet', 'detergent', 'food_butter', 'orchid', 'needle', 'bills', 'stereo', 'window', 'dining_room', 'bench', 'oven', 'bed', 'novel', 'check', 'coffee_filter', 'tray', 'wooden_spoon', 'centerpiece', 'headset', 'conditioner', 'food_steak', 'napkin', 'home_office', 'soap', 'scrabble', 'dirt', 'coffee']\nTask:File documents\nresponse:\n", "output": {"thoughts": "I file all of my paperwork (including bills, receipts, documents, device instructions, etc.) in my black paperwork case that has many different files in order to organize.  I place each piece of paper in its' corresponding subject (as indicated by printed words on each tab).", "actions": ["[WALK] <home_office>", "[WALK] <bills>", "[FIND] <bills>", "[GRAB] <bills>", "[WALK] <filing_cabinet>", "[OPEN] <filing_cabinet>", "[PUTIN] <bills> <filing_cabinet>", "[FIND] <check>", "[GRAB] <check>", "[WALK] <filing_cabinet>", "[PUTIN] <check> <filing_cabinet>", "[FIND] <check>", "[GRAB] <check>", "[WALK] <filing_cabinet>", "[PUTIN] <check> <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'drawing', 'wallshelf', 'floor', 'phone', 'character', 'bedroom', 'alcohol', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'foundation', 'food_noodles', 'wall', 'nightstand', 'bench', 'oven', 'food_food', 'television', 'curtain', 'rag', 'pencil', 'door', 'clothes_scarf', 'home_office', 'table']\nTask:Play games\nresponse:\n{\n    \"thoughts\": I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <board_game>', '[FIND] <board_game>', '[GRAB] <board_game>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <board_game> <table>']\n}\n", "input": "\nScene_Objects:['pot', 'juice', 'tea', 'chair', 'trashcan', 'drawing', 'bills', 'longboard', 'floor', 'man', 'tea_bag', 'food_bacon', 'ceilinglamp', 'laptop', 'shower', 'window', 'foundation', 'couch', 'pillow', 'wall', 'pasta', 'cutting_board', 'basket_for_clothes', 'bed', 'light_bulb', 'television', 'clothes_dress', 'bag', 'home_office', 'table']\nTask:Have conversation with boyfriend\nresponse:\n", "output": {"thoughts": "Sit on sofa, talk", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'bathroom', 'milk', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'purse', 'computer', 'shower', 'window', 'electric_shaver', 'mousepad', 'oil', 'couch', 'form', 'wall', 'food_food', 'television', 'rag', 'lighter', 'mouse', 'controller', 'sink']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": first when i hear the ringing sound i will run to my living room and picks up and i will say hello,\n    \"actions\": ['[RUN] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'cat', 'ceilingfan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'mousepad', 'oil', 'couch', 'wall', 'wall_clock', 'powersocket', 'oven', 'bench', 'newspaper', 'mouthwash', 'television', 'diary', 'towel_rack', 'mouse', 'dresser', 'hairbrush', 'hairdryer', 'soap']\nTask:Check email\nresponse:\n", "output": {"thoughts": "Sit in teak deck chair, have wooden drawing board on lap, laptop on wooden drawing board, log in to my two email accounts, go through new emails and respond.", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[GRAB] <laptop>", "[WALK] <chair>", "[SIT] <chair>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[TOUCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'folder', 'desk', 'alcohol', 'stovefan', 'plate', 'mousepad', 'wall', 'oven', 'food_food', 'check', 'tray', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'sink', 'clothes_scarf', 'light', 'clothes_shirt', 'teeth']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk over and pick up the phone,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'juice', 'microwave', 'toilet', 'chair', 'walllamp', 'dough', 'drawing', 'cpuscreen', 'floor', 'folder', 'desk', 'bookshelf', 'keyboard', 'computer', 'mail', 'shower', 'picture', 'mousepad', 'wall', 'nightstand', 'bench', 'oven', 'photoframe', 'novel', 'check', 'food_food', 'centerpiece', 'mouse']\nTask:Arrange folders\nresponse:\n", "output": {"thoughts": "Go through all my folders and re-arrange in alfabetaical order and put on computer desk", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <folder>", "[GRAB] <folder>", "[OPEN] <desk>", "[PUTBACK] <folder> <desk>", "[FIND] <folder>", "[GRAB] <folder>", "[PUTBACK] <folder> <desk>", "[FIND] <folder>", "[GRAB] <folder>", "[PUTBACK] <folder> <desk>", "[FIND] <folder>", "[GRAB] <folder>", "[PUTBACK] <folder> <desk>", "[FIND] <folder>", "[GRAB] <folder>", "[PUTBACK] <folder> <desk>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'table_cloth', 'mat', 'faucet', 'shower', 'oil', 'clothes_socks', 'dustpan', 'wall', 'toy', 'bed', 'razor', 'water', 'coffee_pot', 'cupboard', 'hairdryer', 'sink', 'napkin', 'bathroom_cabinet', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Wash the teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[TURNTO] <water>', '[POINTAT] <water>', '[FIND] <teeth>', '[WASH] <teeth>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'blender', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'shampoo', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'coin', 'doorjamb', 'mouse', 'sink', 'light', 'soap']\nTask:Write  school paper\nresponse:\n", "output": {"thoughts": "Writing  and drafting a paper for  class work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'walllamp', 'detergent', 'cat', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'remote_control', 'mail', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'couch', 'wall', 'bench', 'clothes_gloves', 'coffe_maker', 'television', 'curtain', 'coffee_pot', 'home_office', 'table', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'chair', 'dog', 'trashcan', 'woman', 'milk', 'man', 'envelope', 'ceilinglamp', 'shower', 'window', 'pillow', 'dining_room', 'newspaper', 'clothes_gloves', 'coffe_maker', 'toilet_paper', 'toothbrush_holder', 'oven_mitts', 'tray', 'curtain', 'console', 'kitchen_counter', 'cupboard', 'maindoor', 'sink', 'napkin', 'home_office', 'tooth_paste']\nTask:Greet people\nresponse:\n", "output": {"thoughts": "Say HI to those that come in.", "actions": ["[WALK] <home_office>", "[WALK] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <woman>", "[GREET] <woman>", "[FIND] <man>", "[GREET] <man>", "[FIND] <man>", "[GREET] <man>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'ceilingfan', 'drawing', 'cpuscreen', 'phone', 'desk', 'keyboard', 'mail', 'mat', 'window', 'towel', 'dining_room', 'wall', 'pasta', 'oven', 'cup', 'check', 'curtain', 'rag', 'doorjamb', 'cards', 'cupboard', 'conditioner', 'clothes_scarf', 'napkin', 'home_office', 'vase', 'soap', 'laser_pointer']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go to my bedroom, and then I switch on the radio,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'chair', 'band-aids', 'orchid', 'hanger', 'floor', 'ceiling', 'phone', 'folder', 'character', 'ceilinglamp', 'shower', 'oil', 'clothes_hat', 'wall', 'cup', 'novel', 'check', 'ice', 'rag', 'doorjamb', 'kettle', 'console', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'dresser', 'cupboard', 'light']\nTask:vacuum carpet\nresponse:\n", "output": {"thoughts": "I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[OPEN] <dresser>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PLUGIN] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[PULL] <vacuum_cleaner>", "[PUSH] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>", "[PLUGOUT] <vacuum_cleaner>", "[WALK] <dresser>", "[PUTIN] <vacuum_cleaner> <dresser>", "[CLOSE] <dresser>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'toilet', 'food_butter', 'orchid', 'bathroom_counter', 'bedroom', 'tea_bag', 'bookshelf', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'window', 'food_noodles', 'couch', 'clothes_hat', 'pillow', 'bench', 'powersocket', 'basket_for_clothes', 'television', 'oven_mitts', 'kitchen_counter', 'shoes', 'pajamas', 'door', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'blender', 'walllamp', 'ceilingfan', 'dough', 'drawing', 'floor', 'ceiling', 'desk', 'stamp', 'remote_control', 'food_cake', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'wall', 'oven', 'photoframe', 'novel', 'television', 'curtain', 'home_office', 'table', 'instrument_guitar']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch nd watch t.V", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'tablelamp', 'mop', 'microwave', 'freezer', 'walllamp', 'needle', 'food_egg', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'stovefan', 'stereo', 'form', 'dining_room', 'wall', 'oven', 'powersocket', 'cutting_board', 'toy', 'keys', 'food_food', 'coffe_maker', 'wine', 'tray', 'kitchen_counter', 'clothes_scarf', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Open the fridge, and put groceries in the fridge. Close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'freezer', 'walllamp', 'food_butter', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'bedroom', 'alcohol', 'window', 'food_jam', 'dining_room', 'wall', 'bench', 'powersocket', 'newspaper', 'food_food', 'doorjamb', 'kettle', 'towel_rack', 'dresser', 'cupboard', 'door', 'table', 'light']\nTask:Eat dinner\nresponse:\n", "output": {"thoughts": "In our dining room, my family eats dinner. Once the food has been prepared, we gather at the dining room table and prepare to eat.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <food_food> <table>", "[WALK] <chair>", "[PULL] <chair>", "[SIT] <chair>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>", "[TURNTO] <food_food>", "[LOOKAT] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'toilet', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'ceiling', 'cleaning_solution', 'mop_bucket', 'table_cloth', 'faucet', 'shower', 'iron', 'clothes_socks', 'pillow', 'wall', 'food_food', 'mirror', 'wine', 'doorjamb', 'coffee_pot', 'paper_towel', 'cupboard', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'dishwasher', 'scrabble']\nTask:Clean mirror\nresponse:\n{\n    \"thoughts\": First spray the mirrors with Windex or window type spray.  Take paper towels and wipe until sparkly clean.  If they have streaks, repeat the process.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[FIND] <cleaning_solution>', '[RINSE] <cleaning_solution>', '[WIPE] <mirror>']\n}\n", "input": "\nScene_Objects:['pot', 'detergent', 'drawing', 'floor', 'ceiling', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'laptop', 'shower', 'window', 'oil', 'clothes_socks', 'couch', 'food_orange', 'pillow', 'wall', 'novel', 'coffe_maker', 'television', 'food_rice', 'rag', 'doorjamb', 'cupboard', 'door', 'sink', 'napkin', 'table', 'light']\nTask:Relax\nresponse:\n", "output": {"thoughts": "Lie down on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'toilet', 'chair', 'freezer', 'detergent', 'food_carrot', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'bedroom', 'piano_bench', 'ceilinglamp', 'mat', 'shower', 'faucet', 'oil', 'food_orange', 'wall', 'nightstand', 'powersocket', 'photoframe', 'food_food', 'check', 'doorjamb', 'vacuum_cleaner', 'door', 'table', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Put the groceries on the table into the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTBACK] <food_chicken> <freezer>', '[CLOSE] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_carrot> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'crayon', 'chair', 'freezer', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'ceilinglamp', 'balanceball', 'mousepad', 'pillow', 'wall', 'nightstand', 'bench', 'razor', 'coin', 'doorjamb', 'mouse', 'conditioner', 'light', 'dishwasher']\nTask:Turn on computer\nresponse:\n", "output": {"thoughts": "Push power button.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'detergent', 'ceilingfan', 'hanger', 'bathroom_counter', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'food_cake', 'clothes_jacket', 'button', 'laptop', 'tvstand', 'mat', 'dustpan', 'food_food', 'after_shave', 'towel_rack', 'pencil', 'kitchen_counter', 'laundry_detergent', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'table', 'vase', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'ceilingfan', 'trashcan', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bookshelf', 'table_cloth', 'mop_bucket', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'food_noodles', 'couch', 'wall', 'nightstand', 'food_food', 'television', 'curtain', 'cupboard', 'table', 'light', 'vase', 'dishwasher', 'dirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit on the couch or the chair or the floor and watch cable, movies, or netflix.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'bathroom', 'cat', 'milk', 'orchid', 'floor', 'dvd_player', 'desk', 'keyboard', 'purse', 'table_cloth', 'mat', 'shower', 'window', 'clothes_socks', 'couch', 'clothes_pants', 'pillow', 'wall', 'powersocket', 'mouthwash', 'television', 'curtain', 'food_onion', 'doorjamb', 'kitchen_counter', 'bathtub', 'napkin', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'hanger', 'floor', 'ceiling', 'phone', 'ceilinglamp', 'tvstand', 'stereo', 'electric_shaver', 'food_noodles', 'form', 'bowl', 'wall', 'nightstand', 'cup', 'check', 'toothbrush_holder', 'after_shave', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'light', 'board_game', 'clothes_shirt']\nTask:Put on coat\nresponse:\n", "output": {"thoughts": "Open closet and grab coat.  Put on coat and fasten the front closed.  Close closet.", "actions": ["[WALK] <home_office>", "[WALK] <dresser>", "[FIND] <dresser>", "[OPEN] <dresser>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTON] <clothes_shirt>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'oil', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'creditcard']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home office. Sit on the chair. Switch on the computer. Type with keyboard for internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'beer', 'toilet', 'freezer', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'shoe-shine_kit', 'faucet', 'couch', 'candle', 'wall', 'nightstand', 'bench', 'powersocket', 'photoframe', 'video_game_controller', 'curtain', 'rag', 'console', 'cupboard', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'table', 'colander']\nTask:Wash sink\nresponse:\n", "output": {"thoughts": "I spray some cleaner and wait a minte or so then wipe it out.", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[GRAB] <cleaning_solution>", "[WALK] <sink>", "[POUR] <cleaning_solution> <sink>", "[PUTOBJBACK] <cleaning_solution>", "[FIND] <rag>", "[GRAB] <rag>", "[WALK] <sink>", "[WIPE] <sink>", "[WIPE] <sink>", "[WIPE] <sink>", "[PUTOBJBACK] <rag>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'mat', 'faucet', 'couch', 'dining_room', 'wall', 'newspaper', 'toilet_paper', 'television', 'oven_mitts', 'water', 'tray', 'vacuum_cleaner', 'controller', 'door', 'sink', 'napkin', 'bathroom_cabinet', 'instrument_guitar', 'light', 'soap', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Wash the teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[TURNTO] <water>', '[POINTAT] <water>', '[FIND] <teeth>', '[WASH] <teeth>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'scissors', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'knifeblock', 'cpuscreen', 'ceiling', 'character', 'stamp', 'bookshelf', 'stovefan', 'plate', 'faucet', 'balanceball', 'mousepad', 'wall', 'cup', 'bed', 'food_food', 'coffe_maker', 'wine', 'kitchen_counter', 'cupboard', 'hair', 'laundry_detergent', 'sink', 'table']\nTask:Get out dish\nresponse:\n", "output": {"thoughts": "Open cupboard.  Get out the dish.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[OPEN] <cupboard>", "[FIND] <plate>", "[GRAB] <plate>", "[CLOSE] <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'walllamp', 'nail_polish', 'dough', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'cards', 'dishrack', 'door', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'blender', 'microwave', 'detergent', 'trashcan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'keyboard', 'toaster', 'laptop', 'stereo', 'window', 'food_peanut_butter', 'dining_room', 'broom', 'nightstand', 'oven', 'cup', 'photoframe', 'bed', 'television', 'after_shave', 'curtain', 'pencil', 'bathroom_cabinet', 'light', 'dishwasher', 'vase']\nTask:Clean\nresponse:\n", "output": {"thoughts": "Clean the welcome mat", "actions": ["[WALK] <home_office>", "[WALK] <vacuum_cleaner>", "[FIND] <vacuum_cleaner>", "[GRAB] <vacuum_cleaner>", "[SWITCHON] <vacuum_cleaner>", "[SWITCHOFF] <vacuum_cleaner>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'glue', 'measuring_cup', 'microwave', 'chair', 'blow_dryer', 'knifeblock', 'drawing', 'orchid', 'floor', 'phone', 'bookshelf', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'wall', 'nightstand', 'photoframe', 'food_food', 'coffee_pot', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'instrument_violin', 'table', 'soap', 'tooth_paste']\nTask:Bring dirty plate to sink\nresponse:\n{\n    \"thoughts\": Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[PUTBACK] <plate> <sink>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'beer', 'chair', 'freezer', 'knifeblock', 'bathroom_counter', 'longboard', 'phone', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'balanceball', 'stereo', 'mousepad', 'clothes_pants', 'nightstand', 'bench', 'bed', 'photoframe', 'novel', 'television', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'food_steak', 'sink', 'hairdryer', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Go to the book shelf, get a book, open the book and read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'beer', 'measuring_cup', 'toilet', 'freezer', 'food_carrot', 'drawing', 'food_egg', 'hanger', 'floor', 'ceiling', 'bedroom', 'mat', 'laptop', 'shower', 'tvstand', 'closetdrawer', 'food_orange', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'bed', 'food_food', 'drinking_glass', 'wine', 'doorjamb', 'door', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Robot Drinks the water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>', '[PUTOBJBACK] <drinking_glass>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['glue', 'sheets', 'food_bread', 'food_pizza', 'tablelamp', 'scissors', 'toilet', 'walllamp', 'woman', 'knifeblock', 'hanger', 'bathroom_counter', 'man', 'tea_bag', 'keyboard', 'mail', 'stereo', 'window', 'couch', 'child', 'powersocket', 'photoframe', 'novel', 'coffe_maker', 'video_game_controller', 'ice', 'rag', 'headset', 'chessboard', 'maindoor']\nTask:Greet guests\nresponse:\n", "output": {"thoughts": "Greet guests when they knock on the door and invite them to enter my home", "actions": ["[WALK] <home_office>", "[WALK] <child>", "[FIND] <child>", "[GREET] <child>", "[FIND] <man>", "[GREET] <man>", "[FIND] <woman>", "[GREET] <woman>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'bathroom', 'orchid', 'food_egg', 'cpuscreen', 'floor', 'phone', 'bedroom', 'desk', 'stovefan', 'toaster', 'ceilinglamp', 'shower', 'mat', 'food_peanut_butter', 'wall', 'pasta', 'powersocket', 'oven', 'cup', 'bed', 'after_shave', 'food_onion', 'chessboard', 'dresser', 'cupboard', 'clothes_scarf', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my living room, find my cat, scrub it using my hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'band-aids', 'drawing', 'longboard', 'floor', 'ceiling', 'folder', 'computer', 'clothes_jacket', 'alcohol', 'ceilinglamp', 'balanceball', 'window', 'foundation', 'couch', 'food_peanut_butter', 'wall', 'powersocket', 'oven', 'television', 'wine', 'towel_rack', 'pencil', 'shoes', 'dresser', 'cupboard', 'table', 'light', 'dishwasher']\nTask:Arrange furniture\nresponse:\n", "output": {"thoughts": "Move furniture where you want it.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[PUSH] <coffee_table>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <couch>", "[PUSH] <couch>", "[WALK] <filing_cabinet>", "[PUSH] <filing_cabinet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'wallshelf', 'bookshelf', 'clothes_jacket', 'toaster', 'ceilinglamp', 'faucet', 'window', 'closetdrawer', 'dining_room', 'nightstand', 'wall', 'wall_clock', 'powersocket', 'basket_for_clothes', 'novel', 'coffe_maker', 'tray', 'doorjamb', 'console', 'dishrack', 'hairdryer', 'door', 'bathtub']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'nail_polish', 'food_carrot', 'drawing', 'knife', 'floor', 'ceiling', 'desk', 'mat', 'stereo', 'shower', 'clothes_socks', 'candle', 'pillow', 'dining_room', 'nightstand', 'bench', 'newspaper', 'check', 'video_game_controller', 'doorjamb', 'cards', 'chessboard', 'pajamas', 'clothes_dress', 'table', 'light', 'board_game', 'laser_pointer', 'coffee']\nTask:Play games\nresponse:\n", "output": {"thoughts": "I either bring out monopoly, uno, chess, checkers or playing cards and lay them out on the dining room table. We then play the games for a couple of hours.", "actions": ["[WALK] <home_office>", "[WALK] <board_game>", "[FIND] <board_game>", "[GRAB] <board_game>", "[WALK] <dining_room>", "[WALK] <table>", "[PUTBACK] <board_game> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'food_butter', 'food_carrot', 'drawing', 'orchid', 'knifeblock', 'knife', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'toaster', 'mat', 'food_noodles', 'couch', 'pillow', 'wall', 'bed', 'photoframe', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'cupboard', 'door', 'bathroom_cabinet', 'table', 'light', 'dirt']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Relax on the sofa in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'drawing', 'orchid', 'floor', 'ceiling', 'bedroom', 'keyboard', 'stovefan', 'ceilinglamp', 'toaster', 'form', 'wall', 'oven', 'food_vegetable', 'photoframe', 'tray', 'kitchen_counter', 'cupboard', 'maindoor', 'home_office', 'dishwasher', 'clothes_skirt', 'soap', 'creditcard']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I prepare meals and cook them on the stove.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[CLOSE] <freezer>", "[FIND] <oven>", "[SWITCHON] <oven>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'ceilinglamp', 'tvstand', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'clothes_gloves', 'novel', 'tray', 'doorjamb', 'kitchen_counter', 'door', 'table', 'light', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['toilet', 'nail_polish', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'toaster', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'cup', 'check', 'television', 'curtain', 'shoes', 'vacuum_cleaner', 'hairdryer', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the living room, I settle on the sofa, I turn on the television, change the channel", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <shoes>", "[FIND] <shoes>", "[GRAB] <shoes>", "[DROP] <shoes>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'microwave', 'chair', 'freezer', 'nail_polish', 'bathroom', 'food_butter', 'knifeblock', 'cpuscreen', 'hanger', 'ceiling', 'dvd_player', 'stamp', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'iron', 'nightstand', 'oven', 'cutting_board', 'check', 'television', 'coin', 'tray', 'doorjamb', 'bathroom_cabinet', 'teeth']\nTask:Answer door\nresponse:\n{\n    \"thoughts\": Walk up to the door.  Open the door and greet whomever is out side,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['crayon', 'cd_player', 'chair', 'toilet', 'walllamp', 'cat', 'hanger', 'floor', 'phone', 'remote_control', 'purse', 'alcohol', 'tvstand', 'iron', 'picture', 'window', 'wall', 'nightstand', 'powersocket', 'bench', 'cutting_board', 'basket_for_clothes', 'after_shave', 'doorjamb', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'door', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.", "actions": ["[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'tablelamp', 'bathroom', 'trashcan', 'food_butter', 'cpuscreen', 'desk', 'bookshelf', 'computer', 'mop_bucket', 'mat', 'closetdrawer', 'form', 'dining_room', 'nightstand', 'bench', 'oven', 'powersocket', 'toy', 'photoframe', 'coffe_maker', 'television', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'bathtub', 'sink', 'home_office']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Grab toys off floor.  Put it in a container.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'piano_bench', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'powersocket', 'doorjamb', 'mouse', 'dresser', 'maindoor', 'light', 'instrument_violin', 'shoe_rack']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Compose an email in the home office.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'blender', 'crayon', 'microwave', 'toilet', 'walllamp', 'milk', 'drawing', 'bathroom_counter', 'phone', 'desk', 'envelope', 'table_cloth', 'ceilinglamp', 'plate', 'faucet', 'picture', 'couch', 'clothes_pants', 'powersocket', 'bench', 'newspaper', 'tray', 'rag', 'kitchen_counter', 'laundry_detergent', 'soap', 'scrabble', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[PLUGIN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'crayon', 'freezer', 'bathroom', 'food_carrot', 'drawing', 'knifeblock', 'floor', 'desk', 'ceilinglamp', 'stereo', 'window', 'couch', 'food_orange', 'pillow', 'form', 'dining_room', 'wall', 'bench', 'coffe_maker', 'television', 'rag', 'pencil', 'comb', 'table', 'tooth_paste', 'clothes_shirt', 'scrabble', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'water_glass', 'pot', 'glue', 'walllamp', 'trashcan', 'knifeblock', 'ceiling', 'clothes_jacket', 'stovefan', 'toaster', 'mat', 'faucet', 'microphone', 'wall', 'cup', 'bed', 'food_food', 'check', 'water', 'food_donut', 'curtain', 'kitchen_counter', 'chessboard', 'cupboard', 'laundry_detergent', 'sink', 'spoon', 'bathroom_cabinet', 'dirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'crayon', 'microwave', 'drawing', 'bills', 'cpuscreen', 'longboard', 'floor', 'bookshelf', 'computer', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'window', 'oil', 'couch', 'food_jam', 'food_orange', 'wall', 'bench', 'mouthwash', 'television', 'cupboard', 'table', 'clothes_skirt', 'light', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'food_ice_cream', 'blow_dryer', 'detergent', 'ceilingfan', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'shoe-shine_kit', 'couch', 'food_orange', 'wall', 'tape', 'television', 'curtain', 'kitchen_counter', 'conditioner', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": For relaxing walk to living room. Find the sofa to relax. Arrange the sofa to relax. Sit on the sofa to relaxing.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[FIND] <conditioner>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'tea', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'food_chicken', 'button', 'stovefan', 'toaster', 'ceilinglamp', 'wall', 'oven', 'bed', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and walk over the cabinent. I pull out a pan and walk to the stove. I put the pan on the stove and walk over to the fridge. I pull out out meat from the fridge and walk back over the stove. Then I put the meat in the pan.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[WALK] <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'television', 'curtain', 'doorjamb', 'mouse', 'door', 'table']\nTask:Watch youtube\nresponse:\n{\n    \"thoughts\": Go to youtube.Com watch videos,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'toilet', 'dough', 'toothbrush', 'drawing', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'stereo', 'shower', 'window', 'oil', 'couch', 'bowl', 'wall', 'oven', 'toy', 'cup', 'bed', 'food_food', 'television', 'wooden_spoon', 'towel_rack', 'comb', 'cupboard', 'table', 'dishwasher']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Find couch. Sit on couch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'microwave', 'chair', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'bedroom', 'ceilinglamp', 'shoe-shine_kit', 'faucet', 'clothes_pants', 'dining_room', 'wall', 'tape', 'basket_for_clothes', 'bed', 'coffe_maker', 'chessboard', 'clothes_dress', 'door', 'home_office', 'light', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'toilet', 'ceilingfan', 'milk', 'knife', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'desk', 'keyboard', 'food_cake', 'computer', 'stovefan', 'deck_of_cards', 'mousepad', 'wall', 'wall_clock', 'photoframe', 'bed', 'video_game_controller', 'coffee_pot', 'mouse', 'door', 'bathroom_cabinet', 'light', 'soap', 'scrabble', 'coffee']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I go to the bathroom. I locate the toilet. I walk to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[WALK] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'detergent', 'food_butter', 'clothes_underwear', 'hanger', 'floor', 'stamp', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'shoe-shine_kit', 'clothes_socks', 'closetdrawer', 'couch', 'clothes_pants', 'basket_for_clothes', 'check', 'mouthwash', 'food_rice', 'vacuum_cleaner', 'laundry_detergent', 'clothes_dress', 'bathroom_cabinet', 'light', 'dishwasher', 'clothes_shirt', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[DROP] <basket_for_clothes>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[GRAB] <laundry_detergent>', '[PUTOBJBACK] <laundry_detergent>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'microwave', 'toilet', 'nail_polish', 'knifeblock', 'hanger', 'bedroom', 'keyboard', 'mail', 'button', 'mat', 'stereo', 'iron', 'shower', 'mousepad', 'dustpan', 'photoframe', 'newspaper', 'food_food', 'video_game_controller', 'coffe_maker', 'coin', 'curtain', 'kettle', 'napkin', 'home_office', 'light', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into the living room and turn on the radio.  Then, I select the station I want, which is usually a preset.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'chair', 'cat', 'drawing', 'floor', 'tea_bag', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'stovefan', 'window', 'couch', 'food_orange', 'wall', 'food_food', 'television', 'oven_mitts', 'food_onion', 'curtain', 'after_shave', 'comb', 'napkin', 'home_office', 'table', 'dishwasher', 'bathroom_cabinet', 'laser_pointer', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room,find the sofa,sat down in the sofa,take some relaxation.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['blender', 'food_ice_cream', 'bathroom', 'cat', 'drawing', 'bills', 'bathroom_counter', 'floor', 'stovefan', 'ceilinglamp', 'mat', 'toaster', 'window', 'mousepad', 'couch', 'towel', 'wall', 'powersocket', 'tape', 'food_cheese', 'novel', 'television', 'wooden_spoon', 'curtain', 'shaving_cream', 'chessboard', 'shoes', 'food_steak', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'blender', 'crayon', 'microwave', 'walllamp', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'bedroom', 'envelope', 'stovefan', 'window', 'oil', 'wall', 'powersocket', 'bench', 'toy', 'keys', 'food_donut', 'wooden_spoon', 'curtain', 'doorjamb', 'maindoor', 'bathtub', 'bathroom_cabinet', 'light', 'dirt', 'teeth']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['pot', 'crayon', 'microwave', 'chair', 'freezer', 'detergent', 'food_egg', 'hanger', 'floor', 'ceiling', 'toaster', 'plate', 'shower', 'mousepad', 'broom', 'wall', 'pasta', 'keys', 'novel', 'coffe_maker', 'toothbrush_holder', 'coffee_filter', 'wooden_spoon', 'towel_rack', 'controller', 'maindoor', 'sink', 'dishwasher', 'coffee', 'creditcard']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[PUTBACK] <coffee> <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'tea', 'knifeblock', 'cpuscreen', 'folder', 'bedroom', 'table_cloth', 'toaster', 'plate', 'homework', 'oil', 'closetdrawer', 'powersocket', 'tape', 'bed', 'novel', 'coffe_maker', 'television', 'razor', 'wine', 'doorjamb', 'shoes', 'dishrack', 'headset', 'cupboard', 'hairdryer', 'light', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk into living room, walk to stereo, turn on stereo, listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'food_ice_cream', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'dvd_player', 'character', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'electric_shaver', 'mousepad', 'couch', 'form', 'wall', 'nightstand', 'clothes_gloves', 'television', 'curtain', 'headset', 'maindoor', 'table', 'clothes_shirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to Living room. Find Television. Walk to sofa. Sit on sofa. Find Remote. Grab remote. Turn to television. Look at television. Switch on Television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_pizza', 'filing_cabinet', 'microwave', 'chair', 'walllamp', 'orchid', 'knifeblock', 'bathroom_counter', 'ceiling', 'phone', 'stamp', 'keyboard', 'thread', 'computer', 'stovefan', 'toaster', 'plate', 'shower', 'mousepad', 'form', 'wall', 'food_cheese', 'bed', 'console', 'headset', 'dresser', 'spoon', 'bathroom_cabinet', 'bag']\nTask:Walk to room\nresponse:\n{\n    \"thoughts\": Leave the living room and enter the hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'filing_cabinet', 'toilet', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'remote_control', 'thread', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'shower', 'couch', 'pillow', 'wall', 'food_cheese', 'television', 'curtain', 'centerpiece', 'home_office', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I go in my living room, find the remote, sit on my couch,  turn on the TV and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'chair', 'freezer', 'detergent', 'bathroom_counter', 'desk', 'bedroom', 'keyboard', 'button', 'stovefan', 'ceilinglamp', 'stereo', 'shower', 'window', 'clothes_socks', 'dining_room', 'nightstand', 'wall', 'bench', 'oven', 'powersocket', 'bed', 'television', 'curtain', 'doorjamb', 'dresser', 'napkin', 'bathroom_cabinet', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room, look for radio, grab on button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'chair', 'walllamp', 'ceilingfan', 'knife', 'orchid', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'balanceball', 'mousepad', 'clothes_socks', 'bowl', 'wall', 'check', 'television', 'mouse', 'kitchen_counter', 'cd', 'door', 'light', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "When the phone rings, I pick up the call and give response to the call.", "actions": ["[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'newspaper', 'television', 'doorjamb', 'dresser', 'soap', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit on my couch and watch TV in my living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'shoe_rack', 'food_salt', 'beer', 'measuring_cup', 'microwave', 'trashcan', 'food_carrot', 'orchid', 'drawing', 'bookshelf', 'envelope', 'alcohol', 'ceilinglamp', 'plate', 'food_jam', 'pillow', 'oven', 'newspaper', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'lighter', 'doorjamb', 'coffee_pot', 'towel_rack', 'cupboard', 'hairdryer', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find a cat, touch it gently", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'sheets', 'food_bread', 'food_pizza', 'tablelamp', 'scissors', 'toilet', 'walllamp', 'woman', 'knifeblock', 'hanger', 'bathroom_counter', 'man', 'tea_bag', 'keyboard', 'mail', 'stereo', 'window', 'couch', 'child', 'powersocket', 'photoframe', 'novel', 'coffe_maker', 'video_game_controller', 'ice', 'rag', 'headset', 'chessboard', 'maindoor']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Greet guests when they knock on the door and invite them to enter my home,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'folder', 'character', 'remote_control', 'food_cake', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'picture', 'window', 'couch', 'wall', 'oven', 'bench', 'photoframe', 'coffee_table', 'food_food', 'television', 'curtain', 'lighter', 'towel_rack', 'controller', 'maindoor', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room and look at the coffee table. Grab the remote control and turn to the television and point the remote control. Put back the remote control.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <coffee_table>", "[TURNTO] <coffee_table>", "[LOOKAT] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'beer', 'chair', 'freezer', 'knifeblock', 'bathroom_counter', 'longboard', 'phone', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'balanceball', 'stereo', 'mousepad', 'clothes_pants', 'nightstand', 'bench', 'bed', 'photoframe', 'novel', 'television', 'pencil', 'kitchen_counter', 'vacuum_cleaner', 'food_steak', 'sink', 'hairdryer', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the book shelf, get a book, open the book and read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['beer', 'toilet', 'walllamp', 'detergent', 'dough', 'drawing', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'dvd_player', 'desk', 'table_cloth', 'faucet', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'pasta', 'photoframe', 'food_food', 'check', 'food_donut', 'dishrack', 'cupboard', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom, find sink, turn on faucet, put soap in hands, wash and rinse hands,", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <soap>", "[WASH] <soap>", "[RINSE] <soap>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'crayon', 'microwave', 'toilet', 'chair', 'dough', 'drawing', 'knifeblock', 'clothes_underwear', 'floor', 'folder', 'bedroom', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'candle', 'clothes_hat', 'wall', 'food_food', 'food_onion', 'cards', 'mouse', 'kitchen_counter', 'cupboard', 'conditioner', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'dirt']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": I walk into the kitchen to the sink. I put dishes in sink and wash them with my hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <sink>', '[PUTBACK] <plate> <sink>', '[WASH] <plate>']\n}\n", "input": "\nScene_Objects:['glue', 'food_pizza', 'tea', 'freezer', 'walllamp', 'knifeblock', 'food_egg', 'cpuscreen', 'bathroom_counter', 'phone', 'computer', 'toaster', 'brush', 'ceilinglamp', 'mat', 'window', 'mousepad', 'pillow', 'wall', 'clothes_gloves', 'novel', 'check', 'food_onion', 'console', 'kitchen_counter', 'vacuum_cleaner', 'hair', 'controller', 'bathroom_cabinet', 'instrument_violin']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to bedroom, sit down, pick up book, read", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'tablelamp', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'toaster', 'stovefan', 'mat', 'faucet', 'shower', 'window', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'water', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'ceilinglamp', 'balanceball', 'shower', 'mousepad', 'pillow', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'cards', 'mouse', 'bathroom_cabinet', 'light', 'soap']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to the home office and turn to the desk. Switch on the computer. Grab the mouse and keyboard to start working.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[TURNTO] <desk>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'dog', 'orchid', 'floor', 'desk', 'button', 'clothes_jacket', 'mat', 'stereo', 'shower', 'window', 'couch', 'pillow', 'wall', 'wall_clock', 'bench', 'cutting_board', 'toilet_paper', 'mouthwash', 'television', 'curtain', 'after_shave', 'doorjamb', 'hair', 'food_steak', 'napkin', 'home_office', 'table', 'vase']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": sitting down to listen to music on the radio,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'beer', 'crayon', 'detergent', 'drawing', 'food_egg', 'floor', 'ceiling', 'bedroom', 'brush', 'mat', 'ceilinglamp', 'shower', 'balanceball', 'window', 'couch', 'towel', 'wall', 'bench', 'light_bulb', 'television', 'water', 'curtain', 'mouse', 'kitchen_counter', 'door', 'home_office', 'soap', 'laser_pointer']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk into bathroom. Turn on shower. Take off clothing. Climb into shower. Rinse with water. Soap up body. Rinse off with water. Turn off shower. Climb out of shower. Dry off with towel.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[TURNTO] <shower>", "[FIND] <water>", "[RINSE] <water>", "[FIND] <soap>", "[WASH] <soap>", "[RINSE] <water>", "[FIND] <towel>", "[GRAB] <towel>", "[SCRUB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'cat', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'phone', 'desk', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'toaster', 'window', 'oil', 'closetdrawer', 'couch', 'form', 'pillow', 'wall', 'wall_clock', 'coffe_maker', 'television', 'curtain', 'mouse', 'home_office', 'table', 'coffee']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <couch>', '[SIT] <couch>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'toilet', 'jelly', 'ground_coffee', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'clothes_jacket', 'faucet', 'mousepad', 'wall', 'oven', 'cup', 'photoframe', 'video_game_controller', 'television', 'tray', 'doorjamb', 'mouse', 'conditioner', 'sink', 'bathroom_cabinet', 'home_office', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom and wash hands.", "actions": ["[RUN] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <sink>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[GRAB] <soap>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'toilet', 'walllamp', 'bathroom', 'detergent', 'drawing', 'bills', 'floor', 'character', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'stereo', 'closetdrawer', 'form', 'dining_room', 'nightstand', 'powersocket', 'cup', 'bed', 'tray', 'mouse', 'kitchen_counter', 'cupboard', 'bathroom_cabinet', 'home_office', 'dirt', 'electrical_outlet']\nTask:Turn on radio\nresponse:\n{\n    \"thoughts\": Go to the radio.  Switch on the power button.  Tune the station.  Set the volume.  Listen to the music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[FIND] <electrical_outlet>', '[TURNTO] <electrical_outlet>', '[PLUGIN] <stereo>', '[SWITCHON] <electrical_outlet>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'television', 'toothbrush_holder', 'curtain', 'kettle', 'mouse', 'bag', 'table', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <chair>", "[SIT] <chair>", "[TURNTO] <television>", "[WATCH] <television>", "[SWITCHOFF] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'microwave', 'toilet', 'drawing', 'wallshelf', 'ceiling', 'desk', 'bookmark', 'toaster', 'stovefan', 'mat', 'shower', 'couch', 'wall', 'oven', 'coffe_maker', 'coin', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'cupboard', 'controller', 'hairdryer', 'spoon', 'dishwasher', 'creditcard']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen, take a glass from the cupboard, fill with water and drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>', '[RINSE] <water_glass>', '[PUTBACK] <water_glass> <cupboard>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'food_ice_cream', 'walllamp', 'nail_polish', 'detergent', 'knife', 'needle', 'cpuscreen', 'bathroom_counter', 'character', 'bedroom', 'remote_control', 'envelope', 'alcohol', 'button', 'deck_of_cards', 'iron', 'nightstand', 'pasta', 'television', 'wine', 'towel_rack', 'mouse', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher', 'vase']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Take the Tv Remote Control  in the Living Room table And Press the Channel Change Button to change the channel", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'chair', 'freezer', 'walllamp', 'blow_dryer', 'toothbrush', 'drawing', 'kitchen_cabinet', 'floor', 'ceiling', 'character', 'bedroom', 'keyboard', 'ceilinglamp', 'stereo', 'shower', 'foundation', 'food_peanut_butter', 'towel', 'wall', 'clothes_gloves', 'food_food', 'toilet_paper', 'tray', 'after_shave', 'dresser', 'cupboard', 'door', 'light']\nTask:Look in refrigerator\nresponse:\n{\n    \"thoughts\": Approach the refrigerator. Use the handle to open the refrigerator door. Look at each shelf of the refrigerator. Look in each refrigerator drawer. Close the refrigerator door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[TURNTO] <freezer>', '[WALK] <freezer>', '[OPEN] <freezer>', '[TURNTO] <freezer>', '[LOOKAT] <freezer>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <food_food>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['juice', 'bathroom', 'ceilingfan', 'drawing', 'bills', 'knife', 'orchid', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'bookmark', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'couch', 'pillow', 'wall', 'oven', 'keys', 'television', 'coffee_filter', 'curtain', 'centerpiece', 'cupboard', 'bathroom_cabinet', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "change the tv channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'folder', 'tea_bag', 'mat', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'pasta', 'nightstand', 'photoframe', 'novel', 'food_food', 'tray', 'curtain', 'doorjamb', 'door', 'table', 'light', 'board_game']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Bring book, lie on the sofa, read book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[LIE] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sponge', 'fork', 'sheets', 'tablelamp', 'mop', 'detergent', 'ceilingfan', 'milk', 'hanger', 'ceiling', 'desk', 'window', 'mousepad', 'clothes_socks', 'clothes_pants', 'pillow', 'dining_room', 'bench', 'cutting_board', 'bed', 'novel', 'check', 'television', 'curtain', 'cards', 'towel_rack', 'cd', 'shoes', 'napkin', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into room, grab phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'filing_cabinet', 'tea', 'microwave', 'dog', 'freezer', 'walllamp', 'drawing', 'orchid', 'hanger', 'keyboard', 'button', 'ceilinglamp', 'stereo', 'homework', 'mousepad', 'closetdrawer', 'food_peanut_butter', 'food_jam', 'wall', 'oven', 'food_cheese', 'keys', 'tray', 'lighter', 'mouse', 'kitchen_counter', 'cd', 'clothes_dress', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Petting the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sponge', 'juice', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'desk', 'mop_bucket', 'stovefan', 'toaster', 'faucet', 'food_jam', 'wall', 'photoframe', 'novel', 'coffe_maker', 'video_game_controller', 'shaving_cream', 'towel_rack', 'hairbrush', 'sink', 'bathroom_cabinet', 'home_office', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash Hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <faucet>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'dog', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'piano_bench', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'pillow', 'towel', 'wall', 'oven', 'coffe_maker', 'television', 'coffee_filter', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <coffee>', '[GRAB] <coffee>', '[PUTBACK] <coffee> <coffe_maker>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'freezer', 'cat', 'orchid', 'drawing', 'hanger', 'ceiling', 'dvd_player', 'desk', 'mail', 'piano_bench', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'food_orange', 'pillow', 'bench', 'bed', 'check', 'doorjamb', 'cd', 'controller', 'laundry_detergent', 'sink', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go to the living room then go to the cat and touch it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'walllamp', 'drawing', 'needle', 'wallshelf', 'floor', 'ceiling', 'shampoo', 'character', 'computer', 'toaster', 'mat', 'tvstand', 'deck_of_cards', 'stovefan', 'ceilinglamp', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'coffe_maker', 'television', 'tray', 'doorjamb', 'pencil', 'conditioner', 'door', 'table', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Find couch. Sit on couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['fork', 'food_bread', 'filing_cabinet', 'tablelamp', 'freezer', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'bills', 'floor', 'ceiling', 'character', 'food_cake', 'piano_bench', 'brush', 'ceilinglamp', 'mousepad', 'towel', 'wall', 'bed', 'food_food', 'food_donut', 'ice', 'curtain', 'kettle', 'dresser', 'light', 'soap', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, put the groceries in the fridge and close it.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'mop', 'cat', 'knife', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'dvd_player', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'faucet', 'stereo', 'window', 'couch', 'towel', 'wall', 'novel', 'tray', 'doorjamb', 'hairbrush', 'sink', 'bathroom_cabinet', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'ceilingfan', 'toothbrush', 'knifeblock', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'piano_bench', 'toaster', 'faucet', 'window', 'mousepad', 'clothes_pants', 'wall', 'nightstand', 'pasta', 'photoframe', 'toothbrush_holder', 'shoes', 'dresser', 'clothes_dress', 'door', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'tooth_paste', 'laser_pointer']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "go to bathroom. grab toothbrush. put toothpaste on toothbrush. brush teeth. turn on faucet. rinse toothbrush and mouth. put toothbrush away.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'glue', 'measuring_cup', 'microwave', 'chair', 'blow_dryer', 'knifeblock', 'drawing', 'orchid', 'floor', 'phone', 'bookshelf', 'toaster', 'ceilinglamp', 'plate', 'faucet', 'wall', 'nightstand', 'photoframe', 'food_food', 'coffee_pot', 'pencil', 'kitchen_counter', 'cupboard', 'door', 'sink', 'instrument_violin', 'table', 'soap', 'tooth_paste']\nTask:Bring dirty plate to sink\nresponse:\n{\n    \"thoughts\": Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[PUTBACK] <plate> <sink>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'filing_cabinet', 'freezer', 'drawing', 'hanger', 'kitchen_cabinet', 'floor', 'phone', 'bathroom_counter', 'folder', 'stamp', 'keyboard', 'computer', 'food_bacon', 'table_cloth', 'brush', 'mat', 'plate', 'clothes_socks', 'couch', 'bowl', 'bench', 'toy', 'bed', 'check', 'cupboard', 'door', 'table', 'light']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the kitchen cabinet. Grab the plate and put it on the table.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'glue', 'measuring_cup', 'dog', 'walllamp', 'bathroom', 'floor', 'phone', 'bedroom', 'mat', 'stereo', 'faucet', 'window', 'closetdrawer', 'dining_room', 'food_cheese', 'standingmirror', 'bed', 'check', 'television', 'coffee_filter', 'food_rice', 'ice', 'curtain', 'doorjamb', 'pencil', 'conditioner', 'bathtub', 'napkin', 'vase']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['scissors', 'tea', 'chair', 'trashcan', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'tea_bag', 'bookshelf', 'keyboard', 'computer', 'picture', 'mousepad', 'food_jam', 'wall', 'cutting_board', 'cup', 'novel', 'tray', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to home office, turn on light, grab a book, sit in chair, start to read the book", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'crayon', 'chair', 'walllamp', 'hanger', 'floor', 'ceiling', 'phone', 'tea_bag', 'clothes_jacket', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'dish_soap', 'mousepad', 'dustpan', 'clothes_hat', 'wall', 'cup', 'coffe_maker', 'coffee_filter', 'curtain', 'coffee_pot', 'mouse', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'napkin']\nTask:Wash dishes\nresponse:\n{\n    \"thoughts\": Place all of the dirty dishes in a sink full of clean, soapy water.  Wash them.  Rinse them in fresh, running water.  Place them in the strainer to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[SWITCHOFF] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <cup>', '[GRAB] <cup>', '[SCRUB] <cup>', '[PUTOBJBACK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[SCRUB] <cup>', '[PUTOBJBACK] <cup>', '[PUTBACK] <sponge> <sink>', '[SWITCHON] <faucet>', '[RINSE] <cup>', '[GRAB] <cup>', '[FIND] <dishrack>', '[PUTBACK] <cup> <dishrack>', '[RINSE] <cup>', '[GRAB] <cup>', '[PUTBACK] <cup> <dishrack>', '[RINSE] <sponge>', '[SWITCHOFF] <faucet>', '[GRAB] <sponge>', '[PUTOBJBACK] <sponge>']\n}\n", "input": "\nScene_Objects:['sheets', 'crayon', 'chair', 'band-aids', 'bathroom', 'food_butter', 'ceilingfan', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'longboard', 'folder', 'bookshelf', 'brush', 'shoe-shine_kit', 'stereo', 'food_jam', 'wall', 'powersocket', 'coffe_maker', 'food_food', 'television', 'wooden_spoon', 'doorjamb', 'dresser', 'cupboard', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk in kitchen. switch on light. walk in living room. switch on light.walk in entrance hall. switch on light.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'milk', 'drawing', 'cpuscreen', 'shampoo', 'folder', 'bookshelf', 'computer', 'table_cloth', 'ceilinglamp', 'mat', 'oil', 'mousepad', 'couch', 'towel', 'nightstand', 'wall', 'cutting_board', 'cup', 'photoframe', 'bed', 'food_donut', 'mouse', 'kitchen_counter', 'pajamas', 'sink', 'napkin', 'home_office', 'tooth_paste']\nTask:Open front door\nresponse:\n{\n    \"thoughts\": Walk over to the front door, turn the knob, and open the door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'mop', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'envelope', 'faucet', 'picture', 'homework', 'window', 'mousepad', 'clothes_hat', 'towel', 'wall', 'photoframe', 'hands_both', 'food_food', 'doorjamb', 'kettle', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'light', 'vase', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I need to rinse my dentures off and put them back in my mouth.", "actions": ["[WALK] <bathroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <teeth>", "[GRAB] <teeth>", "[WASH] <teeth>", "[PUTOBJBACK] <teeth>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <towel>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'chair', 'jelly', 'knifeblock', 'floor', 'ceiling', 'character', 'alcohol', 'toaster', 'plate', 'faucet', 'drying_rack', 'window', 'tvstand', 'dish_soap', 'closetdrawer', 'couch', 'towel', 'pillow', 'food_kiwi', 'wall', 'toy', 'novel', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'instrument_guitar', 'shoe_rack']\nTask:Do dishes\nresponse:\n{\n    \"thoughts\": Wash dirty dishes, dry them and then put them away,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sink>', '[PUTOBJBACK] <dish_soap>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[FIND] <drying_rack>', '[PUTBACK] <plate> <drying_rack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <drying_rack>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['sponge', 'tea', 'toilet', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'ceiling', 'bookshelf', 'computer', 'button', 'mail', 'stovefan', 'ceilinglamp', 'window', 'couch', 'wall', 'bench', 'tape', 'cup', 'coffe_maker', 'tray', 'headset', 'comb', 'cupboard', 'sink', 'home_office', 'table', 'bathroom_cabinet', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up your book and read it", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'bedroom', 'keyboard', 'tvstand', 'stereo', 'balanceball', 'picture', 'mousepad', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'photoframe', 'bed', 'toilet_paper', 'coffe_maker', 'cards', 'mouse', 'kitchen_counter', 'cd', 'vase']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['food_bread', 'freezer', 'cat', 'food_carrot', 'dough', 'orchid', 'needle', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'ceilinglamp', 'mat', 'stereo', 'couch', 'dustpan', 'pillow', 'wall', 'powersocket', 'light_bulb', 'food_food', 'coffe_maker', 'video_game_controller', 'oven_mitts', 'food_rice', 'vacuum_cleaner', 'dresser', 'door', 'hairdryer']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I leave the bags with groceries near the fridge, open the door and put the groceries in the shelves.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'toaster', 'mat', 'faucet', 'shower', 'pillow', 'wall', 'nightstand', 'bench', 'toy', 'coffe_maker', 'television', 'curtain', 'after_shave', 'doorjamb', 'towel_rack', 'kitchen_counter', 'chessboard', 'sink', 'home_office', 'bathroom_cabinet', 'table', 'instrument_violin']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'food_ice_cream', 'mop', 'chair', 'freezer', 'walllamp', 'cat', 'trashcan', 'toothbrush', 'hanger', 'bathroom_counter', 'floor', 'keyboard', 'dining_room', 'bench', 'powersocket', 'cup', 'photoframe', 'novel', 'toothbrush_holder', 'curtain', 'mouse', 'headset', 'pajamas', 'door', 'dishwasher', 'vase', 'tooth_paste', 'clothes_shirt', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, grab the toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now wash the teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'detergent', 'cat', 'food_butter', 'knife', 'phone', 'character', 'table_cloth', 'mat', 'deck_of_cards', 'stereo', 'form', 'pillow', 'dining_room', 'nightstand', 'wall', 'basket_for_clothes', 'photoframe', 'tray', 'doorjamb', 'kitchen_counter', 'chessboard', 'dresser', 'clothes_dress', 'table', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'mop', 'dog', 'jelly', 'drawing', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'character', 'computer', 'stovefan', 'faucet', 'oil', 'dining_room', 'wall', 'nightstand', 'oven', 'cutting_board', 'photoframe', 'check', 'tray', 'sink', 'napkin', 'bathroom_cabinet', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "go to the bathroom and wash your hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[SQUEEZE] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'scissors', 'knife', 'hanger', 'bathroom_counter', 'floor', 'bedroom', 'desk', 'keyboard', 'piano_bench', 'stovefan', 'stereo', 'window', 'wall', 'powersocket', 'cutting_board', 'cup', 'bed', 'console', 'pencil', 'shoes', 'door', 'bathtub', 'bag', 'instrument_guitar', 'table', 'home_office', 'light', 'board_game']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into my room and turn the radio on so that I can listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['food_bread', 'microwave', 'freezer', 'walllamp', 'bathroom', 'toothbrush', 'orchid', 'hanger', 'longboard', 'phone', 'desk', 'computer', 'food_bacon', 'table_cloth', 'clothes_jacket', 'brush', 'ceilinglamp', 'faucet', 'foundation', 'mousepad', 'clothes_pants', 'pasta', 'light_bulb', 'toothbrush_holder', 'coin', 'coffee_filter', 'towel_rack', 'hair', 'vase', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "WALK TO THE BATHROOM USE PASTE AND BRUSH TO WASH YOUR TEETH", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'tablelamp', 'microwave', 'detergent', 'clothes_underwear', 'ceiling', 'character', 'desk', 'mail', 'table_cloth', 'bookmark', 'shoe-shine_kit', 'faucet', 'electric_shaver', 'mousepad', 'candle', 'food_jam', 'food_orange', 'nightstand', 'powersocket', 'newspaper', 'coin', 'food_onion', 'pencil', 'dresser', 'home_office', 'dishwasher', 'tooth_paste', 'coffee']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['water_glass', 'scissors', 'tablelamp', 'bathroom', 'drawing', 'floor', 'desk', 'computer', 'food_bacon', 'stovefan', 'ceilinglamp', 'shower', 'window', 'couch', 'clothes_hat', 'towel', 'wall', 'oven', 'photoframe', 'check', 'food_food', 'television', 'coffe_maker', 'tray', 'centerpiece', 'pencil', 'chessboard', 'hair', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Drink a glass of water in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'tea', 'microwave', 'chair', 'walllamp', 'blow_dryer', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'plate', 'mousepad', 'food_noodles', 'couch', 'closetdrawer', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'razor', 'television', 'shaving_cream', 'kitchen_counter', 'bathroom_cabinet', 'dishwasher', 'soap', 'teeth']\nTask:Text friends while sitting on couch\nresponse:\n{\n    \"thoughts\": Text while sitting,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[TOUCH] <phone>']\n}\n", "input": "\nScene_Objects:['fork', 'blow_dryer', 'bathroom', 'ceilingfan', 'drawing', 'hanger', 'kitchen_cabinet', 'floor', 'phone', 'desk', 'computer', 'laptop', 'plate', 'stereo', 'oil', 'clothes_hat', 'wall', 'bench', 'oven', 'powersocket', 'newspaper', 'coffe_maker', 'curtain', 'rag', 'kitchen_counter', 'maindoor', 'home_office', 'table', 'light', 'soap']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Grab plates and forks. Walk into dining room. Put plates and forks on table.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[PUTBACK] <fork> <table>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'chair', 'drawing', 'bills', 'knife', 'floor', 'ceilinglamp', 'stereo', 'window', 'couch', 'food_jam', 'form', 'towel', 'dining_room', 'wall', 'basket_for_clothes', 'photoframe', 'novel', 'food_food', 'check', 'television', 'oven_mitts', 'razor', 'tray', 'pencil', 'dresser', 'pajamas', 'table', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit and read,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[WALK] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['chair', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'tea_bag', 'keyboard', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'stovefan', 'window', 'mat', 'plate', 'couch', 'candle', 'microphone', 'dining_room', 'wall', 'powersocket', 'box', 'photoframe', 'coffe_maker', 'television', 'curtain', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Turn on the TV, sit on the couch, find a show, and watch it.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ironing_board', 'sheets', 'tablelamp', 'detergent', 'knifeblock', 'bathroom_counter', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'iron', 'window', 'couch', 'pillow', 'bowl', 'nightstand', 'food_cheese', 'video_game_controller', 'tray', 'curtain', 'doorjamb', 'towel_rack', 'kitchen_counter', 'pajamas', 'bag', 'dishwasher', 'board_game', 'clothes_shirt']\nTask:Iron shirt\nresponse:\n{\n    \"thoughts\": Set up an ironing board and an iron.  Lay a shirt on board.  Run iron over shirt until it is free of wrinkles.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <ironing_board>', '[FIND] <ironing_board>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <ironing_board>', '[FIND] <iron>', '[GRAB] <iron>', '[PLUGIN] <iron>', '[PLUGOUT] <iron>', '[PUTOBJBACK] <iron>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'tablelamp', 'blender', 'scissors', 'freezer', 'walllamp', 'ground_coffee', 'orchid', 'character', 'bedroom', 'bookshelf', 'alcohol', 'button', 'toaster', 'stereo', 'homework', 'electric_shaver', 'mousepad', 'couch', 'bowl', 'toy', 'clothes_gloves', 'coffe_maker', 'kitchen_counter', 'conditioner', 'napkin', 'table', 'light', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room, look for radio, grab on button", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'keyboard', 'mail', 'mat', 'faucet', 'wall', 'nightstand', 'powersocket', 'oven', 'cup', 'photoframe', 'bed', 'novel', 'curtain', 'doorjamb', 'diary', 'cupboard', 'door', 'sink', 'instrument_guitar']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk over to the bookshelf. Pick storybook off the shelf. Sit on bed. Open storybook and read out loud.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[WALK] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'ground_coffee', 'walllamp', 'detergent', 'knifeblock', 'longboard', 'floor', 'ceiling', 'toaster', 'stovefan', 'plate', 'faucet', 'candle', 'wall', 'box', 'coffe_maker', 'food_food', 'television', 'coffee_filter', 'water', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'hair', 'maindoor', 'sink', 'napkin', 'dishwasher', 'tooth_paste']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the coffee machine. I open the coffee machine and pour a filter in it. I open the cupboard and grab ground coffee. I put ground coffee in the filter and water in the coffee machine. I close the coffee machine. I switch on the coffee machine and wait", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'clothes_socks', 'closetdrawer', 'mousepad', 'clothes_pants', 'wall', 'powersocket', 'oven', 'cup', 'newspaper', 'food_food', 'check', 'coffe_maker', 'shaving_cream', 'mouse', 'pencil', 'shoes', 'dresser', 'hair', 'conditioner', 'light', 'clothes_shirt', 'shoe_rack']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Open drawer. Pull out pants and put them on. Take a shirt out and put it on. Put on socks and shoes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTON] <clothes_socks>', '[CLOSE] <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'bathroom', 'food_carrot', 'toothbrush', 'orchid', 'knife', 'longboard', 'character', 'bedroom', 'computer', 'faucet', 'homework', 'clothes_socks', 'food_noodles', 'dining_room', 'photoframe', 'check', 'television', 'after_shave', 'rag', 'lighter', 'towel_rack', 'kitchen_counter', 'dresser', 'clothes_dress', 'home_office', 'light', 'laser_pointer', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office. I find the telephone. I pick up the telephone handset.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'mop', 'band-aids', 'walllamp', 'trashcan', 'ceilingfan', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'desk', 'computer', 'stovefan', 'toaster', 'deck_of_cards', 'shower', 'wall', 'pasta', 'oven', 'cup', 'water', 'tray', 'doorjamb', 'shaving_cream', 'kitchen_counter', 'sauce_pan', 'cupboard', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Cook some pasta on the stove,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pasta>', '[GRAB] <pasta>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <sauce_pan>', '[POUR] <water> <sauce_pan>', '[FIND] <oven>', '[SWITCHON] <oven>', '[PUTBACK] <pasta> <sauce_pan>', '[FIND] <sauce_pan>', '[GRAB] <sauce_pan>', '[PUTBACK] <sauce_pan> <oven>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'freezer', 'cat', 'knifeblock', 'longboard', 'floor', 'ceiling', 'keyboard', 'clothes_jacket', 'stovefan', 'ceilinglamp', 'mat', 'stereo', 'tvstand', 'electric_shaver', 'mousepad', 'couch', 'wall', 'oven', 'bench', 'cup', 'food_food', 'doorjamb', 'conditioner', 'clothes_scarf', 'home_office', 'table', 'light', 'vase']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I get my groceries and go to the fridge. I open the fridge and put the groceries there.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'crayon', 'dog', 'toothbrush', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'faucet', 'shower', 'balanceball', 'mousepad', 'wall', 'pasta', 'powersocket', 'oven', 'nightstand', 'photoframe', 'video_game_controller', 'coffee_filter', 'wooden_spoon', 'console', 'shoes', 'sink', 'bathroom_cabinet', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": Put toothpaste on toothbrush, brush teeth, rinse mouth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>', '[PUTOBJBACK] <tooth_paste>', '[WALK] <sink>', '[FIND] <teeth>', '[SCRUB] <teeth>', '[SCRUB] <teeth>', '[SCRUB] <teeth>', '[PUTBACK] <toothbrush> <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <toothbrush>', '[RINSE] <teeth>', '[SWITCHOFF] <faucet>', '[GRAB] <toothbrush>', '[PUTOBJBACK] <toothbrush>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'chair', 'trashcan', 'ceilingfan', 'bills', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'stamp', 'bookshelf', 'envelope', 'computer', 'stovefan', 'pillow', 'wall', 'powersocket', 'coffe_maker', 'television', 'oven_mitts', 'ice', 'doorjamb', 'towel_rack', 'dresser', 'cupboard', 'door', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Coming home turning on lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'walllamp', 'bathroom', 'toothbrush', 'drawing', 'orchid', 'floor', 'desk', 'computer', 'ceilinglamp', 'mat', 'deck_of_cards', 'stereo', 'window', 'couch', 'food_jam', 'pillow', 'wall', 'light_bulb', 'food_donut', 'tray', 'curtain', 'shoes', 'dishrack', 'sink', 'spoon', 'table', 'dishwasher', 'coffee']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'scissors', 'measuring_cup', 'toilet', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'remote_control', 'purse', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'form', 'dining_room', 'wall', 'powersocket', 'television', 'curtain', 'food_onion', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room. Find remote control. Look at remote control. Push power on button. Find sofa. Sit down on sofa. Watch Television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'tablelamp', 'walllamp', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'ceilinglamp', 'floor_lamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'broom', 'wall', 'oven', 'photoframe', 'novel', 'television', 'doorjamb', 'dresser', 'cupboard', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bookshelf', 'computer', 'stovefan', 'toaster', 'food_noodles', 'wall', 'oven', 'toy', 'food_vegetable', 'food_food', 'video_game_controller', 'water', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'dishwasher', 'scrabble']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "walk to kicthen , find stove , find cooking pot , find vegetables , find water  ,put vegetables in cooking pot, pour water to cooking pot,put cooking pot on stove, switch on stove", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[FIND] <pot>", "[FIND] <food_vegetable>", "[FIND] <water>", "[GRAB] <water>", "[WASH] <food_vegetable>", "[POUR] <water> <pot>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'detergent', 'food_butter', 'dough', 'food_carrot', 'drawing', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'keyboard', 'computer', 'clothes_jacket', 'tvstand', 'faucet', 'shower', 'electric_shaver', 'mousepad', 'towel', 'wall', 'oven', 'cup', 'photoframe', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap and hot water,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'cpuscreen', 'clothes_underwear', 'floor', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'mail', 'toaster', 'laptop', 'balanceball', 'window', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'photoframe', 'novel', 'coffe_maker', 'coffee_filter', 'mouse', 'clothes_skirt', 'vase']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I open my laptop and then login to Windows. Then I open Google Chrome, and type gmail.com into the address bar. My email automatically logs in, then I click the Compose button, type in the address of the recipient, then the subject of the email, and then start typing my email", "actions": ["[WALK] <laptop>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_bread', 'toilet', 'bathroom', 'detergent', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'mail', 'ceilinglamp', 'shower', 'iron', 'closetdrawer', 'couch', 'towel', 'oven', 'coin', 'lighter', 'console', 'mouse', 'shoes', 'cd', 'dishrack', 'spoon', 'light', 'table']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'crayon', 'toilet', 'walllamp', 'milk', 'orchid', 'knife', 'ceiling', 'dvd_player', 'character', 'desk', 'thread', 'mop_bucket', 'toaster', 'oil', 'food_noodles', 'candle', 'bowl', 'bench', 'powersocket', 'check', 'mouthwash', 'television', 'maindoor', 'spoon', 'home_office', 'bathroom_cabinet', 'light', 'vase', 'laser_pointer']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the radio. Turn on the radio. Listen to the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'walllamp', 'trashcan', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'stovefan', 'faucet', 'food_jam', 'wall', 'nightstand', 'oven', 'cup', 'photoframe', 'light_bulb', 'bed', 'clothes_gloves', 'television', 'water', 'tray', 'lighter', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'colander']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, grab a cup, fill with water and then drink it.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[FIND] <sink>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'chair', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'window', 'mousepad', 'couch', 'wall', 'powersocket', 'bed', 'curtain', 'doorjamb', 'shaving_cream', 'mouse', 'comb', 'door', 'napkin', 'table', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I turn on computer. I use keyboard to type. I open browser. I search.", "actions": ["[WALK] <bedroom>", "[WALK] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'microwave', 'drawing', 'orchid', 'wallshelf', 'floor', 'table_cloth', 'toaster', 'ceilinglamp', 'plate', 'shower', 'faucet', 'dish_soap', 'closetdrawer', 'pillow', 'wall', 'novel', 'coffe_maker', 'food_rice', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'clothes_scarf', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find dishwasher. Open dishwasher. Find dishrack. Find plate 1. Grab plate 1. Put plate 1 in dishrack. Repeat for dish 2. Find dish soap. Grab dish soap. Put dish soap to dishwasher. Close dishwasher. Switch on dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'blow_dryer', 'cat', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'ceilinglamp', 'balanceball', 'shower', 'window', 'couch', 'wall', 'powersocket', 'food_food', 'television', 'oven_mitts', 'doorjamb', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'light', 'table', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "My cat looks like it wants attention, so I will go pet it on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'freezer', 'walllamp', 'drawing', 'orchid', 'needle', 'floor', 'phone', 'ceiling', 'character', 'keyboard', 'food_cake', 'computer', 'stovefan', 'mat', 'plate', 'window', 'closetdrawer', 'wall', 'powersocket', 'tape', 'television', 'tray', 'curtain', 'towel_rack', 'sink', 'spoon', 'bathtub', 'table', 'bathroom_cabinet']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Place your plate on the placemat. The forks will be placed on the left side of the plate and the knife and spoon on the right and place the spoon to the right of the knife. If you want to set out food, place them above the plate.,\n    \"actions\": ['[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[DROP] <plate>', '[FIND] <fork>', '[GRAB] <fork>', '[DROP] <fork>', '[FIND] <spoon>', '[GRAB] <spoon>', '[DROP] <spoon>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'dog', 'bathroom', 'food_carrot', 'bills', 'ceiling', 'phone', 'keyboard', 'stovefan', 'stereo', 'window', 'oven', 'bench', 'tape', 'bed', 'light_bulb', 'food_onion', 'coffee_pot', 'kettle', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'maindoor', 'instrument_guitar', 'table', 'bathroom_cabinet', 'dishwasher', 'coffee']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my living room, find my cat, scrub it using my hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'tea_bag', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'hair', 'door']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": I walk to the Home Office then find the computer and sit in the chair. I switch on the computer and start typing the email. I switch the computer off.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'filing_cabinet', 'juice', 'freezer', 'cat', 'drawing', 'knife', 'floor', 'ceiling', 'shampoo', 'desk', 'clothes_jacket', 'button', 'ceilinglamp', 'mat', 'balanceball', 'window', 'couch', 'clothes_pants', 'wall', 'oven', 'tape', 'photoframe', 'television', 'tray', 'curtain', 'table', 'soap', 'shoe_rack']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I see my cat on the couch so I walk over, sit down and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'clothes_underwear', 'floor', 'ceiling', 'thread', 'computer', 'bookmark', 'toaster', 'stovefan', 'couch', 'wall', 'powersocket', 'cup', 'food_food', 'razor', 'water', 'tray', 'centerpiece', 'pencil', 'kitchen_counter', 'chessboard', 'cupboard', 'clothes_scarf', 'light', 'table', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['fork', 'scissors', 'tea', 'freezer', 'walllamp', 'knifeblock', 'knife', 'bills', 'floor', 'toaster', 'mat', 'plate', 'stereo', 'picture', 'tvstand', 'dish_soap', 'microphone', 'wall', 'oven', 'cup', 'light_bulb', 'coffe_maker', 'drinking_glass', 'tray', 'mouse', 'cupboard', 'controller', 'maindoor', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and turn to the dishwasher, put the knife, glass, fork and plate into the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[TURNTO] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'tablelamp', 'blender', 'beer', 'toilet', 'jelly', 'freezer', 'blow_dryer', 'detergent', 'orchid', 'knifeblock', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'mail', 'computer', 'bookmark', 'stereo', 'window', 'couch', 'microphone', 'check', 'television', 'rag', 'pencil', 'controller', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Pet Cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'ground_coffee', 'trashcan', 'desk', 'bedroom', 'purse', 'mop_bucket', 'shower', 'picture', 'homework', 'mousepad', 'form', 'wall', 'oven', 'bed', 'keys', 'video_game_controller', 'wine', 'food_onion', 'after_shave', 'console', 'mouse', 'cd', 'comb', 'dresser', 'door', 'clothes_skirt', 'dirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat; pet the cat by touching it", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'freezer', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'button', 'bookmark', 'ceilinglamp', 'window', 'couch', 'microphone', 'pillow', 'wall', 'pasta', 'tape', 'box', 'novel', 'curtain', 'doorjamb', 'hairbrush', 'door', 'home_office', 'table', 'bathroom_cabinet', 'soap', 'tooth_paste', 'clothes_shirt']\nTask:Text friends while sitting on couch\nresponse:\n{\n    \"thoughts\": Text while sitting,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[TOUCH] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_carrot', 'drawing', 'floor', 'character', 'computer', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'candle', 'wall', 'wall_clock', 'oven', 'cup', 'photoframe', 'bed', 'novel', 'check', 'toilet_paper', 'television', 'curtain', 'towel_rack', 'mouse', 'hairbrush', 'laundry_detergent', 'table', 'clothes_skirt']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to the couch in the living room, sit on ti and read the book", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'console', 'dresser', 'cupboard', 'home_office', 'table', 'clothes_skirt', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'toilet', 'jelly', 'drawing', 'orchid', 'cpuscreen', 'floor', 'purse', 'ceilinglamp', 'mat', 'shower', 'window', 'couch', 'food_orange', 'dining_room', 'wall', 'box', 'food_cheese', 'photoframe', 'bed', 'coffe_maker', 'television', 'tray', 'doorjamb', 'centerpiece', 'towel_rack', 'cd', 'table', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I cross the room to the couch. I sit down and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'beer', 'toilet', 'band-aids', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'desk', 'mail', 'computer', 'ceilinglamp', 'shoe-shine_kit', 'stereo', 'iron', 'shower', 'pillow', 'nightstand', 'toy', 'bed', 'food_food', 'oven_mitts', 'coffee_filter', 'after_shave', 'curtain', 'lighter', 'cupboard', 'maindoor']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the cat. I pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'freezer', 'ceilingfan', 'food_carrot', 'knifeblock', 'clothes_underwear', 'floor', 'ceiling', 'bedroom', 'desk', 'stamp', 'alcohol', 'food_bacon', 'table_cloth', 'ceilinglamp', 'mousepad', 'food_peanut_butter', 'pillow', 'wall', 'food_food', 'check', 'television', 'towel_rack', 'kettle', 'kitchen_counter', 'dishrack', 'hair', 'dishwasher', 'vase']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'drawing', 'orchid', 'floor', 'folder', 'mop_bucket', 'ceilinglamp', 'stereo', 'picture', 'window', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'bench', 'powersocket', 'novel', 'coffe_maker', 'toilet_paper', 'toothbrush_holder', 'tray', 'curtain', 'lighter', 'headset', 'cupboard', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and find the sofa. I sit on the sofa and get comfortable.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'freezer', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'balanceball', 'picture', 'window', 'clothes_hat', 'broom', 'wall', 'pasta', 'novel', 'doorjamb', 'coffee_pot', 'vacuum_cleaner', 'dresser', 'cupboard', 'clothes_dress', 'door', 'maindoor', 'bag', 'light', 'table', 'vase']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk over and pick up the phone", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'drawing', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'tvstand', 'shower', 'closetdrawer', 'wall', 'wall_clock', 'oven', 'cup', 'standingmirror', 'check', 'television', 'centerpiece', 'console', 'mouse', 'vacuum_cleaner', 'shoes', 'comb', 'dresser', 'hairbrush', 'sink', 'bag', 'napkin', 'bathroom_cabinet', 'vase']\nTask:vacuum carpet\nresponse:\n{\n    \"thoughts\": I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'freezer', 'walllamp', 'detergent', 'food_butter', 'longboard', 'floor', 'bathroom_counter', 'bedroom', 'keyboard', 'toaster', 'plate', 'dish_soap', 'bowl', 'broom', 'wall', 'pasta', 'oven', 'bed', 'coffe_maker', 'check', 'food_food', 'coffee_filter', 'tray', 'doorjamb', 'headset', 'cupboard', 'light', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'crayon', 'bathroom', 'drawing', 'floor', 'phone', 'character', 'keyboard', 'remote_control', 'button', 'ceilinglamp', 'shower', 'window', 'foundation', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'check', 'food_food', 'television', 'dresser', 'sink', 'table', 'dishwasher', 'clothes_skirt', 'tooth_paste', 'dirt']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": go to living room. grab remote control. sit on couch. point remote at tv. push buttons.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'freezer', 'walllamp', 'bathroom', 'ground_coffee', 'cat', 'blow_dryer', 'ceilingfan', 'bathroom_counter', 'bedroom', 'desk', 'clothes_jacket', 'toaster', 'ceilinglamp', 'tvstand', 'stereo', 'light_bulb', 'coffe_maker', 'mouthwash', 'kitchen_counter', 'chessboard', 'cupboard', 'clothes_dress', 'door', 'clothes_scarf', 'table', 'dishwasher', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "walk to kitchen. grab clothes_dress. look at washing machine. open washing machine. put clothes_dress on washing machine. close washing machine. switch on washing machine. switch off washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <clothes_dress>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[TURNTO] <washing_machine>", "[LOOKAT] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>", "[SWITCHOFF] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'microwave', 'chair', 'ground_coffee', 'walllamp', 'detergent', 'trashcan', 'drawing', 'knife', 'floor', 'ceiling', 'thread', 'toaster', 'dustpan', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'chessboard', 'cupboard', 'spoon', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find coffee maker. Find coffee filter. Put it in the coffee maker. Find ground coffee. Put it in the coffee maker. Find Water. Put it in the coffe maker. Switch on the coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[FIND] <water>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sponge', 'juice', 'toilet', 'walllamp', 'toothbrush', 'drawing', 'cpuscreen', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'stamp', 'faucet', 'balanceball', 'clothes_pants', 'wall', 'powersocket', 'photoframe', 'novel', 'toothbrush_holder', 'television', 'food_donut', 'tray', 'towel_rack', 'cupboard', 'controller', 'sink', 'bathroom_cabinet', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I go into the bathroom, go to the sink, get toothbrush and toothpaste. Put toothpaste on the toothbrush, brush teeth, turn on faucet, rinse brush, and return toothbrush.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <toothbrush>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'band-aids', 'drawing', 'orchid', 'kitchen_cabinet', 'floor', 'bathroom_counter', 'phone', 'bookshelf', 'computer', 'stovefan', 'mat', 'plate', 'ceilinglamp', 'tvstand', 'window', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'bench', 'powersocket', 'oven', 'cutting_board', 'curtain', 'mouse', 'table', 'dishwasher', 'light']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the kitchen cabinet. Grab the plate and put it on the table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'drawing', 'orchid', 'clothes_underwear', 'hanger', 'bathroom_counter', 'dvd_player', 'desk', 'bookshelf', 'ceilinglamp', 'mat', 'plate', 'shower', 'homework', 'candle', 'towel', 'pillow', 'bed', 'food_food', 'television', 'food_donut', 'pencil', 'kitchen_counter', 'shoes', 'hair', 'sink', 'laser_pointer', 'dirt', 'coffee']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "go to living room, find phone, pick up phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'food_salt', 'food_butter', 'food_carrot', 'drawing', 'food_egg', 'floor', 'phone', 'ceiling', 'character', 'purse', 'mat', 'plate', 'window', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'bed', 'photoframe', 'tray', 'coffee_pot', 'kitchen_counter', 'shoes', 'pajamas', 'sink', 'spoon', 'bathtub', 'table', 'vase']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Place your plate on the placemat. The forks will be placed on the left side of the plate and the knife and spoon on the right and place the spoon to the right of the knife. If you want to set out food, place them above the plate.,\n    \"actions\": ['[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[DROP] <plate>', '[FIND] <fork>', '[GRAB] <fork>', '[DROP] <fork>', '[FIND] <spoon>', '[GRAB] <spoon>', '[DROP] <spoon>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'walllamp', 'trashcan', 'food_carrot', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'food_chicken', 'desk', 'stovefan', 'ceilinglamp', 'toaster', 'balanceball', 'shower', 'faucet', 'food_noodles', 'bowl', 'wall', 'oven', 'powersocket', 'cup', 'bed', 'food_food', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'blender', 'food_salt', 'toilet', 'walllamp', 'bathroom', 'food_butter', 'drawing', 'hanger', 'shampoo', 'bedroom', 'purse', 'computer', 'stovefan', 'laptop', 'ceilinglamp', 'electric_shaver', 'window', 'homework', 'clothes_socks', 'wall_clock', 'shelf', 'basket_for_clothes', 'bed', 'coffe_maker', 'check', 'wooden_spoon', 'centerpiece', 'lighter', 'kettle']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": The task is browse and surf on the internet regarding a particular topic,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <laptop>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[WATCH] <laptop>']\n}\n", "input": "\nScene_Objects:['freezer', 'drawing', 'cpuscreen', 'floor', 'stamp', 'desk', 'keyboard', 'bookshelf', 'brush', 'ceilinglamp', 'laptop', 'balanceball', 'toaster', 'window', 'couch', 'clothes_pants', 'broom', 'wall', 'pasta', 'oven', 'nightstand', 'box', 'mouthwash', 'television', 'razor', 'curtain', 'pajamas', 'sink', 'table', 'clothes_skirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I sit in the sofa and I relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'broom', 'wall', 'pasta', 'powersocket', 'shelf', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'light', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find the computer and turn it on. I look at the computer screen. I move and click the mouse to select buttons on the screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[PUSH] <mouse>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'filing_cabinet', 'crayon', 'walllamp', 'bathroom', 'toothbrush', 'floor', 'food_cake', 'toaster', 'plate', 'shower', 'faucet', 'electric_shaver', 'dish_soap', 'broom', 'wall', 'oven', 'bench', 'photoframe', 'light_bulb', 'novel', 'coffe_maker', 'television', 'tray', 'towel_rack', 'kettle', 'dishrack', 'table', 'dishwasher', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'mousepad', 'closetdrawer', 'pillow', 'bowl', 'wall', 'powersocket', 'cup', 'novel', 'coffe_maker', 'food_donut', 'doorjamb', 'mouse', 'dresser', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find a desk and grab a book. Walk to the chair, sit in the chair and read a book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['fork', 'pot', 'filing_cabinet', 'beer', 'bathroom', 'detergent', 'trashcan', 'cpuscreen', 'ceiling', 'bedroom', 'desk', 'keyboard', 'clothes_jacket', 'tvstand', 'couch', 'pillow', 'food_kiwi', 'oven', 'bench', 'coffe_maker', 'oven_mitts', 'towel_rack', 'pencil', 'cupboard', 'hair', 'clothes_dress', 'food_steak', 'napkin', 'vase', 'laser_pointer']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the bedroom. I walk to stereo. I turn on stereo.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'glue', 'filing_cabinet', 'microwave', 'jelly', 'freezer', 'ceilingfan', 'orchid', 'food_egg', 'floor', 'phone', 'ceiling', 'desk', 'thread', 'mat', 'laptop', 'deck_of_cards', 'faucet', 'foundation', 'pillow', 'nightstand', 'powersocket', 'basket_for_clothes', 'food_food', 'lighter', 'sink', 'home_office', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'detergent', 'hanger', 'floor', 'phone', 'stamp', 'toaster', 'balanceball', 'oil', 'pillow', 'dining_room', 'wall_clock', 'toy', 'newspaper', 'keys', 'photoframe', 'food_food', 'clothes_gloves', 'rag', 'doorjamb', 'mouse', 'dishrack', 'dresser', 'cupboard', 'hairdryer', 'spoon', 'table', 'light', 'scrabble']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Pet Cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_pizza', 'juice', 'toilet', 'freezer', 'blow_dryer', 'walllamp', 'ceilingfan', 'knifeblock', 'bathroom_counter', 'alcohol', 'button', 'stereo', 'form', 'towel', 'oven', 'basket_for_clothes', 'photoframe', 'light_bulb', 'toilet_paper', 'toothbrush_holder', 'kitchen_counter', 'shoes', 'chessboard', 'vacuum_cleaner', 'clothes_scarf', 'home_office', 'dishwasher', 'soap', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and get the soap and wash my both hand using the soap and rinse my hands and use towel.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[FIND] <towel>', '[RINSE] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'form', 'pillow', 'wall', 'oven', 'box', 'newspaper', 'coffe_maker', 'television', 'oven_mitts', 'curtain', 'doorjamb', 'pencil', 'cupboard', 'maindoor', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to living room. Walk to and sit at the sofa. Grab remote control and switch on TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'mop', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'envelope', 'faucet', 'picture', 'homework', 'window', 'mousepad', 'clothes_hat', 'towel', 'wall', 'photoframe', 'hands_both', 'food_food', 'doorjamb', 'kettle', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'light', 'vase', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": I need to rinse my dentures off and put them back in my mouth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <teeth>', '[GRAB] <teeth>', '[WASH] <teeth>', '[PUTOBJBACK] <teeth>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <towel>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'fork', 'tablelamp', 'walllamp', 'cpuscreen', 'bathroom_counter', 'floor', 'toaster', 'plate', 'tvstand', 'dish_soap', 'food_noodles', 'bowl', 'wall', 'pasta', 'oven', 'box', 'coffe_maker', 'television', 'ice', 'tray', 'curtain', 'console', 'towel_rack', 'pencil', 'maindoor', 'sink', 'instrument_guitar', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Wash a plate using the dishwasher in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'freezer', 'ceilingfan', 'clothes_underwear', 'bathroom_counter', 'phone', 'keyboard', 'computer', 'table_cloth', 'mop_bucket', 'tvstand', 'mousepad', 'candle', 'clothes_pants', 'basket_for_clothes', 'cup', 'bed', 'food_food', 'curtain', 'kettle', 'mouse', 'kitchen_counter', 'shoes', 'controller', 'clothes_dress', 'clothes_scarf', 'light', 'soap', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I need to get the dirty clothes basket.  Walk to bathroom.  Put all the clothes in washing machine.  Add soap.  Turn on machine.,\n    \"actions\": ['[WALK] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'cat', 'ceilingfan', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'dvd_player', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'foundation', 'oil', 'couch', 'towel', 'wall', 'pasta', 'powersocket', 'television', 'curtain', 'doorjamb', 'shoes', 'pajamas', 'maindoor', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'measuring_cup', 'microwave', 'dog', 'orchid', 'wallshelf', 'floor', 'bedroom', 'computer', 'mop_bucket', 'ceilinglamp', 'faucet', 'electric_shaver', 'couch', 'food_peanut_butter', 'broom', 'cutting_board', 'photoframe', 'newspaper', 'mouthwash', 'ice', 'tray', 'curtain', 'kitchen_counter', 'shoes', 'headset', 'bathroom_cabinet', 'board_game']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my living room, find my cat, scrub it using my hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[SCRUB] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'mop', 'freezer', 'food_carrot', 'knifeblock', 'drawing', 'cpuscreen', 'phone', 'envelope', 'computer', 'ceilinglamp', 'tvstand', 'deck_of_cards', 'foundation', 'candle', 'dining_room', 'bench', 'toy', 'cup', 'bed', 'tray', 'wooden_spoon', 'doorjamb', 'kitchen_counter', 'cupboard', 'clothes_scarf', 'bathroom_cabinet', 'light', 'table', 'home_office']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the entrance hall. Find the telephone. Grab the telephone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'microwave', 'cd_player', 'walllamp', 'drawing', 'orchid', 'bills', 'floor', 'ceiling', 'desk', 'tea_bag', 'button', 'piano_bench', 'stovefan', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'ice', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'door', 'bathtub', 'home_office', 'table', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to living room, walk to cd player, turn on cd player, press play button, walk to sofa, sit in sofa and listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cd_player>', '[SWITCHON] <cd_player>', '[FIND] <button>', '[PUSH] <button>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'food_salt', 'microwave', 'ceilingfan', 'cpuscreen', 'hanger', 'bathroom_counter', 'character', 'keyboard', 'toaster', 'mat', 'balanceball', 'stereo', 'window', 'microphone', 'bench', 'basket_for_clothes', 'clothes_gloves', 'check', 'oven_mitts', 'kettle', 'dresser', 'cupboard', 'maindoor', 'napkin', 'light', 'clothes_skirt', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Wash the clothes in the washing machine", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'mop', 'ground_coffee', 'cat', 'floor', 'ceiling', 'envelope', 'stovefan', 'mat', 'tvstand', 'stereo', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'cutting_board', 'check', 'coffe_maker', 'television', 'video_game_controller', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'chessboard', 'dresser', 'door', 'table', 'colander']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'microwave', 'bathroom', 'trashcan', 'toothbrush', 'drawing', 'knifeblock', 'food_egg', 'floor', 'ceiling', 'desk', 'purse', 'food_bacon', 'ceilinglamp', 'plate', 'laptop', 'deck_of_cards', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'photoframe', 'check', 'television', 'spoon', 'home_office', 'table', 'teeth']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the sofa, and lie down on the sofa.", "actions": ["[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'measuring_cup', 'walllamp', 'knifeblock', 'knife', 'drawing', 'floor', 'ceiling', 'character', 'toaster', 'mat', 'plate', 'shower', 'dish_soap', 'clothes_socks', 'wall', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'dishrack', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet', 'instrument_guitar', 'dishwasher', 'bathtub']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <dishrack>', '[PUTBACK] <fork> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTBACK] <spoon> <dishrack>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishrack>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_ice_cream', 'beer', 'chair', 'freezer', 'blow_dryer', 'bathroom', 'detergent', 'trashcan', 'longboard', 'floor', 'ceiling', 'phone', 'bookshelf', 'ceilinglamp', 'stereo', 'homework', 'oil', 'wall', 'cup', 'food_food', 'check', 'television', 'coffee_filter', 'curtain', 'diary', 'cupboard', 'conditioner', 'sink', 'bag']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk to the kitchen, open the fridge and put the groceries in the fridge. Then I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'ground_coffee', 'walllamp', 'milk', 'cpuscreen', 'ceiling', 'phone', 'keyboard', 'stovefan', 'plate', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'wine', 'tray', 'food_onion', 'mouse', 'kitchen_counter', 'headset', 'comb', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Get out dish\nresponse:\n{\n    \"thoughts\": Open cupboard.  Get out the dish.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[OPEN] <cupboard>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <cupboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'detergent', 'ceilingfan', 'hanger', 'bathroom_counter', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'food_cake', 'clothes_jacket', 'button', 'laptop', 'tvstand', 'mat', 'dustpan', 'food_food', 'after_shave', 'towel_rack', 'pencil', 'kitchen_counter', 'laundry_detergent', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'table', 'vase', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "\"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\"", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[TURNTO] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tea', 'ground_coffee', 'walllamp', 'hanger', 'longboard', 'floor', 'ceiling', 'button', 'toaster', 'faucet', 'mousepad', 'candle', 'wall', 'nightstand', 'powersocket', 'keys', 'novel', 'coffe_maker', 'clothes_gloves', 'coffee_filter', 'water', 'ice', 'tray', 'food_onion', 'cupboard', 'door', 'light', 'dishwasher', 'soap']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['mop', 'microwave', 'toilet', 'band-aids', 'food_butter', 'ceilingfan', 'character', 'bedroom', 'desk', 'stovefan', 'form', 'dining_room', 'bench', 'cutting_board', 'basket_for_clothes', 'clothes_gloves', 'check', 'water', 'tray', 'curtain', 'towel_rack', 'coffee_pot', 'mouse', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'door', 'vase', 'soap', 'dirt']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "put soap on hands, scrub it and rinse with water", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'chair', 'bills', 'orchid', 'drawing', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'bookshelf', 'brush', 'ceilinglamp', 'tvstand', 'stereo', 'stovefan', 'mat', 'mousepad', 'wall', 'powersocket', 'bed', 'television', 'doorjamb', 'dresser', 'clothes_scarf', 'light', 'dishwasher', 'table', 'colander', 'clothes_shirt']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk to the Living Room. Switch on 1 light. Walk to dining room. Switch on 1 light. Walk to Entrance hall. Switch on 1 light.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <dining_room>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'crayon', 'microwave', 'toilet', 'dog', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'desk', 'computer', 'toaster', 'plate', 'faucet', 'deck_of_cards', 'dish_soap', 'form', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'toy', 'photoframe', 'towel_rack', 'cupboard', 'sink', 'bathroom_cabinet']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Make some dish water, grab a dish rag, put dishes in dish water and wash.", "actions": ["[RUN] <sink>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[PUTBACK] <dish_soap> <plate>", "[RINSE] <dish_soap>", "[FIND] <plate>", "[WASH] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'hanger', 'floor', 'ceiling', 'phone', 'ceilinglamp', 'tvstand', 'stereo', 'electric_shaver', 'food_noodles', 'form', 'bowl', 'wall', 'nightstand', 'cup', 'check', 'toothbrush_holder', 'after_shave', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'light', 'board_game', 'clothes_shirt']\nTask:Put on coat\nresponse:\n{\n    \"thoughts\": Open closet and grab coat.  Put on coat and fasten the front closed.  Close closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'drawing', 'keyboard', 'bookmark', 'stovefan', 'ceilinglamp', 'stereo', 'shower', 'foundation', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'bed', 'check', 'television', 'wine', 'curtain', 'towel_rack', 'vacuum_cleaner', 'maindoor', 'door', 'dishwasher', 'vase']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter living room. Walk to telephone. Pick up from hook.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'filing_cabinet', 'microwave', 'chair', 'bathroom', 'orchid', 'character', 'desk', 'bedroom', 'thread', 'computer', 'brush', 'ceilinglamp', 'faucet', 'stereo', 'closetdrawer', 'dining_room', 'wall', 'pasta', 'toy', 'cup', 'standingmirror', 'television', 'wine', 'doorjamb', 'kitchen_counter', 'conditioner', 'bathroom_cabinet', 'clothes_shirt']\nTask:Open door\nresponse:\n{\n    \"thoughts\": Open the door when leaving or coming in.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'food_pizza', 'bathroom', 'trashcan', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bookshelf', 'alcohol', 'mop_bucket', 'ceilinglamp', 'window', 'electric_shaver', 'couch', 'food_kiwi', 'wall', 'broom', 'nightstand', 'toilet_paper', 'television', 'tray', 'mouse', 'dishrack', 'cupboard', 'home_office', 'table', 'instrument_violin']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "lie in the sofa, relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_oatmeal', 'microwave', 'freezer', 'bathroom', 'nail_polish', 'dough', 'food_carrot', 'orchid', 'drawing', 'knife', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'brush', 'mat', 'couch', 'pillow', 'food_cereal', 'wall', 'food_vegetable', 'food_food', 'television', 'console', 'cards', 'kitchen_counter', 'cupboard', 'hair', 'door', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_vegetable> <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTIN] <food_carrot> <freezer>', '[FIND] <food_cereal>', '[GRAB] <food_cereal>', '[PUTIN] <food_cereal> <freezer>', '[FIND] <food_oatmeal>', '[GRAB] <food_oatmeal>', '[PUTIN] <food_oatmeal> <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[FIND] <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sponge', 'spectacles', 'filing_cabinet', 'blender', 'ceilingfan', 'drawing', 'orchid', 'floor', 'purse', 'computer', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'window', 'clothes_socks', 'couch', 'food_jam', 'wall', 'pasta', 'bench', 'television', 'wine', 'mouse', 'pencil', 'dresser', 'door', 'table', 'dishwasher', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to living room. Sit on sofa. Lie down on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'ceilingfan', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'food_kiwi', 'wall', 'oven', 'powersocket', 'cup', 'light_bulb', 'novel', 'television', 'tray', 'curtain', 'after_shave', 'kitchen_counter', 'dresser', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_pizza', 'pot', 'microwave', 'bathroom', 'bills', 'knifeblock', 'cpuscreen', 'drawing', 'longboard', 'floor', 'phone', 'folder', 'bedroom', 'keyboard', 'mail', 'food_bacon', 'piano_bench', 'button', 'table_cloth', 'shower', 'couch', 'dining_room', 'wall', 'powersocket', 'cutting_board', 'water', 'tray', 'pencil', 'door']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drinking a glass of water", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'chair', 'toothbrush', 'drawing', 'floor', 'ceiling', 'computer', 'toaster', 'stovefan', 'mat', 'shower', 'faucet', 'couch', 'wall', 'oven', 'cup', 'coffe_maker', 'food_food', 'mouthwash', 'drinking_glass', 'wine', 'tray', 'towel_rack', 'kitchen_counter', 'cupboard', 'pajamas', 'sink', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'ground_coffee', 'bathroom', 'detergent', 'bathroom_counter', 'bedroom', 'bookshelf', 'button', 'toaster', 'mat', 'ceilinglamp', 'window', 'mousepad', 'couch', 'dining_room', 'photoframe', 'bed', 'novel', 'coffe_maker', 'food_food', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'sink', 'dishwasher', 'clothes_skirt', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room , find phone ,pick up phone 3", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'microwave', 'band-aids', 'drawing', 'cpuscreen', 'hanger', 'floor', 'computer', 'ceilinglamp', 'balanceball', 'window', 'oil', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'coffe_maker', 'toilet_paper', 'television', 'kitchen_counter', 'chessboard', 'home_office', 'table', 'bathroom_cabinet', 'vase', 'light', 'coffee']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Sitting down and relaxing on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['crayon', 'microwave', 'blow_dryer', 'cat', 'milk', 'bathroom_counter', 'phone', 'food_cake', 'computer', 'stovefan', 'balanceball', 'homework', 'couch', 'clothes_hat', 'bench', 'oven', 'powersocket', 'cup', 'newspaper', 'coffe_maker', 'tray', 'wooden_spoon', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'cd', 'cupboard', 'laundry_detergent', 'door']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting a cat may sound simple.Letting the Cat Come to You.Extend a hand or finger and allow the cat a chance to touch her nose to you. When a cat bumps her head into your hand, it's a signal she wants attention.Focusing on Areas with Scent Glands. Feed with food.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <milk>", "[FIND] <milk>", "[GRAB] <milk>", "[POUR] <milk> <cup>", "[TURNTO] <cat>", "[POINTAT] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'freezer', 'walllamp', 'detergent', 'bills', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'shower', 'faucet', 'closetdrawer', 'microphone', 'dining_room', 'pasta', 'mouthwash', 'mouse', 'pencil', 'dresser', 'pajamas', 'controller', 'sink', 'napkin', 'home_office', 'dishwasher', 'soap']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'walllamp', 'ceilingfan', 'drawing', 'knifeblock', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'clothes_jacket', 'stovefan', 'tvstand', 'faucet', 'mousepad', 'food_jam', 'broom', 'wall', 'nightstand', 'clothes_gloves', 'food_food', 'razor', 'water', 'kitchen_counter', 'cupboard', 'door', 'sink', 'spoon', 'napkin']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen, take a glass from the cupboard, fill with water and drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>", "[RINSE] <water_glass>", "[PUTBACK] <water_glass> <cupboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'needle', 'cpuscreen', 'floor', 'character', 'keyboard', 'table_cloth', 'plate', 'faucet', 'window', 'oil', 'mousepad', 'pillow', 'wall', 'oven', 'food_cheese', 'bed', 'tray', 'kettle', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'light']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[TURNTO] <plate>', '[LOOKAT] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WASH] <plate>', '[PUTBACK] <plate> <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'walllamp', 'bathroom', 'cat', 'ceilingfan', 'knifeblock', 'phone', 'character', 'brush', 'toaster', 'mat', 'faucet', 'ceilinglamp', 'shower', 'window', 'couch', 'wall_clock', 'cup', 'newspaper', 'food_onion', 'lighter', 'towel_rack', 'mouse', 'kitchen_counter', 'headset', 'cupboard', 'table', 'vase', 'laser_pointer']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Every time this animal is seen, you must approach it and show compassion for it by petting it in a gentle manner.", "actions": ["[WALK] <home_office>", "[WALK] <bedroom>", "[WALK] <bedroom>", "[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[FIND] <brush>", "[GRAB] <brush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'doorjamb', 'towel_rack', 'pencil', 'shoes', 'dresser', 'sink', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go to the the living room, sit down on the sofa, and then I switch on the TV with the TV remote,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'freezer', 'walllamp', 'detergent', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'stamp', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'couch', 'food_peanut_butter', 'form', 'pillow', 'wall', 'bench', 'novel', 'food_food', 'oven_mitts', 'food_rice', 'diary', 'towel_rack', 'controller', 'maindoor', 'light', 'clothes_skirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find groceries. Grab groceries. Find fridge. Walk to fridge. Put groceries in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_bread', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'stovefan', 'mat', 'tvstand', 'window', 'closetdrawer', 'microphone', 'food_orange', 'pillow', 'wall', 'nightstand', 'bed', 'food_rice', 'curtain', 'kettle', 'dresser', 'pajamas', 'bathtub', 'bathroom_cabinet', 'table', 'vase']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <sheets>', '[PULL] <sheets>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'photoframe', 'food_food', 'doorjamb', 'mouse', 'shoes', 'door', 'home_office', 'light', 'dishwasher']\nTask:Work\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the computer and turn it on. I look at the computer screen. I move and click the mouse to select buttons on the screen.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[FIND] <mouse>", "[PUSH] <mouse>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'pillow', 'wall', 'nightstand', 'bench', 'food_food', 'doorjamb', 'kettle', 'mouse', 'pencil', 'headset', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": go to the office, turn computer on, open an browser or any application to open any website.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'detergent', 'orchid', 'food_egg', 'hanger', 'bathroom_counter', 'floor', 'phone', 'brush', 'tvstand', 'foundation', 'couch', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'cutting_board', 'food_cheese', 'coffe_maker', 'food_food', 'toothbrush_holder', 'wine', 'tray', 'mouse', 'laundry_detergent', 'home_office', 'table', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[PLUGIN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'detergent', 'food_butter', 'clothes_underwear', 'hanger', 'floor', 'stamp', 'bookshelf', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'shoe-shine_kit', 'clothes_socks', 'closetdrawer', 'couch', 'clothes_pants', 'basket_for_clothes', 'check', 'mouthwash', 'food_rice', 'vacuum_cleaner', 'laundry_detergent', 'clothes_dress', 'bathroom_cabinet', 'light', 'dishwasher', 'clothes_shirt', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[DROP] <basket_for_clothes>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <clothes_socks>', '[GRAB] <clothes_socks>', '[PUTBACK] <clothes_socks> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[GRAB] <laundry_detergent>', '[PUTOBJBACK] <laundry_detergent>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['food_bread', 'scissors', 'chair', 'bathroom', 'detergent', 'cat', 'ceilingfan', 'dough', 'knifeblock', 'hanger', 'floor', 'ceiling', 'character', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'microphone', 'wall', 'oven', 'coffe_maker', 'tray', 'curtain', 'lighter', 'kitchen_counter', 'vacuum_cleaner', 'dresser', 'home_office', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find a cat in the kitchen and touch it", "actions": ["[WALK] <dining_room>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'ground_coffee', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mail', 'bookmark', 'toaster', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'mousepad', 'dining_room', 'wall', 'nightstand', 'oven', 'cup', 'bed', 'television', 'coffee_filter', 'mouse', 'door']\nTask:Answer emails\nresponse:\n{\n    \"thoughts\": Open email application.  Open new emails.  Respond accordingly,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <mail>', '[GRAB] <mail>', '[READ] <mail>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[SWITCHOFF] <laptop>']\n}\n", "input": "\nScene_Objects:['food_salt', 'beer', 'measuring_cup', 'freezer', 'ground_coffee', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'envelope', 'computer', 'toaster', 'stereo', 'wall', 'photoframe', 'light_bulb', 'novel', 'coffe_maker', 'coffee_filter', 'water', 'wine', 'curtain', 'doorjamb', 'coffee_pot', 'dresser', 'dishwasher', 'soap']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Making Coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <water>", "[GRAB] <water>", "[OPEN] <coffe_maker>", "[POUR] <water> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <coffee_pot>", "[GRAB] <coffee_pot>", "[PUTBACK] <coffee_pot> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'detergent', 'trashcan', 'woman', 'knifeblock', 'needle', 'cpuscreen', 'desk', 'man', 'keyboard', 'bedroom', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'nightstand', 'oven', 'cup', 'food_food', 'check', 'doorjamb', 'mouse', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'light']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Answer the door. Smile and give a warm hello, or other appropriate greeting. With newcomers to your home, offer your hand to be shook. With old friends it is appropriate to hug, or kiss one another on the cheek.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>', '[FIND] <woman>', '[GREET] <woman>']\n}\n", "input": "\nScene_Objects:['glue', 'crayon', 'microwave', 'food_butter', 'drawing', 'knifeblock', 'floor', 'bedroom', 'envelope', 'stovefan', 'ceilinglamp', 'mat', 'balanceball', 'window', 'couch', 'clothes_hat', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'television', 'after_shave', 'cards', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'hairdryer', 'spoon', 'table', 'light']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Walk to the sofa and sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'jelly', 'chair', 'freezer', 'bathroom', 'food_carrot', 'knifeblock', 'bathroom_counter', 'keyboard', 'alcohol', 'clothes_jacket', 'stovefan', 'mat', 'toaster', 'stereo', 'picture', 'ceilinglamp', 'dining_room', 'bowl', 'wall', 'oven', 'powersocket', 'cup', 'photoframe', 'check', 'video_game_controller', 'razor', 'food_food', 'curtain', 'shoes']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": turn radio on, listen to music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['fork', 'food_ice_cream', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'wall', 'wall_clock', 'basket_for_clothes', 'bed', 'food_food', 'television', 'curtain', 'hairbrush', 'controller', 'bathroom_cabinet', 'table', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room, walk to TV, grab TV remote, turn on TV, put down the remote, walk to sofa, sit in soft, watch TV", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'band-aids', 'bathroom', 'cat', 'orchid', 'wallshelf', 'hanger', 'floor', 'phone', 'character', 'bedroom', 'purse', 'computer', 'table_cloth', 'tvstand', 'powersocket', 'toy', 'clothes_gloves', 'coffe_maker', 'razor', 'food_onion', 'centerpiece', 'headset', 'chessboard', 'cupboard', 'door', 'home_office', 'table', 'laser_pointer']\nTask:Open door\nresponse:\n{\n    \"thoughts\": I walk up to the door.  I use my hand to turn the door knob to the left.  I pull the door open with the same hand I used to turn the door knob,\n    \"actions\": ['[RUN] <home_office>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'piano_bench', 'table_cloth', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'nightstand', 'oven', 'mouthwash', 'television', 'wine', 'curtain', 'doorjamb', 'mouse', 'dishrack', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to the living room to watch TV", "actions": ["[RUN] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[TOUCH] <remote_control>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>", "[DROP] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'ceilingfan', 'trashcan', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'character', 'bookshelf', 'table_cloth', 'mop_bucket', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'food_noodles', 'couch', 'wall', 'nightstand', 'food_food', 'television', 'curtain', 'cupboard', 'table', 'light', 'vase', 'dishwasher', 'dirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on the couch or the chair or the floor and watch cable, movies, or netflix.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['blender', 'food_salt', 'beer', 'cpuscreen', 'desk', 'toaster', 'ceilinglamp', 'stovefan', 'tvstand', 'faucet', 'homework', 'floor_lamp', 'window', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'cutting_board', 'cup', 'bed', 'novel', 'coffe_maker', 'food_rice', 'ice', 'dresser', 'cupboard', 'laundry_detergent', 'bathroom_cabinet', 'tooth_paste', 'shoe_rack']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Turn a light on and read a book in the bedroom.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'food_salt', 'chair', 'walllamp', 'detergent', 'drawing', 'clothes_underwear', 'floor', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'food_noodles', 'garbage_can', 'mousepad', 'microphone', 'wall', 'nightstand', 'food_vegetable', 'bed', 'food_food', 'coffee_filter', 'mouse', 'sink', 'home_office', 'table', 'dishwasher', 'pantry']\nTask:Organize pantry\nresponse:\n{\n    \"thoughts\": Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <pantry>', '[FIND] <food_food>', '[GRAB] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[FIND] <garbage_can>', '[PUTBACK] <food_food> <garbage_can>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <dry_pasta>', '[GRAB] <dry_pasta>', '[TURNTO] <dry_pasta>', '[LOOKAT] <dry_pasta>', '[PUTOBJBACK] <dry_pasta>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[PUSH] <food_food>', '[FIND] <food_noodles>', '[GRAB] <food_noodles>', '[TURNTO] <food_noodles>', '[LOOKAT] <food_noodles>', '[PUTOBJBACK] <food_noodles>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'dog', 'blow_dryer', 'toothbrush', 'orchid', 'clothes_underwear', 'hanger', 'bathroom_counter', 'ceiling', 'purse', 'food_bacon', 'toaster', 'ceilinglamp', 'shower', 'clothes_socks', 'mousepad', 'oil', 'clothes_pants', 'microphone', 'pillow', 'pajamas', 'controller', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'dishwasher', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I grab the basket for clothes from the bedroom. I carry it and walk to the washing machine. I open the washing machine. I grab the dirty clothes. I put the clothes in the washing machine. I grab the laundry detergent and put it in the washing machine. I close the washing machine. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[DROP] <basket_for_clothes>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[GRAB] <laundry_detergent>", "[PUTOBJBACK] <laundry_detergent>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'beer', 'chair', 'walllamp', 'bathroom', 'cat', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'tooth_paste', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'coin', 'doorjamb', 'dresser', 'board_game']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch TV in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'dough', 'ceiling', 'character', 'computer', 'stovefan', 'toaster', 'tvstand', 'oil', 'pillow', 'pasta', 'powersocket', 'cup', 'bed', 'novel', 'food_food', 'mouthwash', 'curtain', 'rag', 'towel_rack', 'cupboard', 'controller', 'door', 'spoon', 'home_office', 'table']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into the living room. I find the stereo and turn it on, and I listen to the music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'ground_coffee', 'bathroom', 'detergent', 'bathroom_counter', 'bedroom', 'bookshelf', 'button', 'toaster', 'mat', 'ceilinglamp', 'window', 'mousepad', 'couch', 'dining_room', 'photoframe', 'bed', 'novel', 'coffe_maker', 'food_food', 'mouse', 'pencil', 'kitchen_counter', 'dresser', 'sink', 'dishwasher', 'clothes_skirt', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to living room , find phone ,pick up phone 3,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'phone', 'tea_bag', 'computer', 'button', 'bookmark', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'couch', 'clothes_hat', 'wall', 'oven', 'bench', 'television', 'curtain', 'vacuum_cleaner', 'dresser', 'maindoor', 'spoon', 'home_office', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room, find the television, switch on the television, find the sofa, walk to sofa, sit in sofa, watch the television", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'blow_dryer', 'cat', 'toothbrush', 'orchid', 'floor', 'keyboard', 'mat', 'ceilinglamp', 'shoe-shine_kit', 'picture', 'window', 'mousepad', 'couch', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'shelf', 'photoframe', 'television', 'ice', 'curtain', 'after_shave', 'doorjamb', 'diary', 'vacuum_cleaner', 'conditioner', 'home_office', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I would walk into the living room. Then sit on the couch and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'walllamp', 'ceilingfan', 'food_carrot', 'toothbrush', 'drawing', 'desk', 'bookshelf', 'stovefan', 'balanceball', 'shower', 'window', 'candle', 'clothes_pants', 'food_orange', 'dining_room', 'bench', 'bed', 'keys', 'food_food', 'diary', 'mouse', 'shoes', 'sink', 'vase', 'tooth_paste', 'scrabble', 'teeth']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "EARLY MORNING AND BEFORE GOING TO BED", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'window', 'toaster', 'clothes_pants', 'broom', 'powersocket', 'bench', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'dresser', 'hair', 'door', 'sink', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Get some clothes together that make an appropriate outfit.  Put them on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['jelly', 'trashcan', 'food_carrot', 'drawing', 'food_egg', 'bathroom_counter', 'ceiling', 'keyboard', 'bookshelf', 'mop_bucket', 'tvstand', 'iron', 'oil', 'food_noodles', 'mousepad', 'dustpan', 'towel', 'nightstand', 'cup', 'light_bulb', 'razor', 'food_donut', 'rag', 'shaving_cream', 'towel_rack', 'pencil', 'kitchen_counter', 'bag', 'bathroom_cabinet', 'dishwasher']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'freezer', 'bathroom', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'bookshelf', 'shower', 'homework', 'wall', 'powersocket', 'box', 'bed', 'coffe_maker', 'television', 'after_shave', 'doorjamb', 'towel_rack', 'mouse', 'cards', 'cupboard', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into living room. Find wall. Find light button. Switch on light button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'measuring_cup', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'doorjamb', 'mouse', 'pencil', 'cupboard', 'door', 'maindoor', 'home_office', 'light', 'dishwasher', 'creditcard']\nTask:Work\nresponse:\n", "output": {"thoughts": "turn computer on, open application to work with", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'toaster', 'mat', 'ceilinglamp', 'plate', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'bench', 'photoframe', 'coffe_maker', 'food_rice', 'dresser', 'cupboard', 'bathroom_cabinet', 'table', 'dishwasher', 'home_office']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Get dishes. Place dishes neatly around table. Get food. Place food on middle of table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <plate> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>', '[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[PUTBACK] <food_food> <table>']\n}\n", "input": "\nScene_Objects:['sheets', 'juice', 'band-aids', 'walllamp', 'detergent', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'faucet', 'shower', 'tvstand', 'balanceball', 'dustpan', 'towel', 'wall', 'photoframe', 'television', 'curtain', 'doorjamb', 'dishrack', 'door', 'sink', 'bathroom_cabinet', 'shoe_rack']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Open the Bathroom Cabinet. Grab a Towel. Close the Bathroom Cabinet. Rinse the Shower.", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <towel>", "[GRAB] <towel>", "[CLOSE] <bathroom_cabinet>", "[FIND] <shower>", "[RINSE] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'microwave', 'freezer', 'drawing', 'floor', 'ceiling', 'dvd_player', 'keyboard', 'stovefan', 'mat', 'laptop', 'stereo', 'oil', 'mousepad', 'couch', 'pillow', 'wall', 'food_food', 'television', 'drinking_glass', 'water', 'curtain', 'diary', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'light', 'instrument_violin']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to the kitchen,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <water>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['band-aids', 'walllamp', 'floor', 'keyboard', 'toaster', 'mat', 'plate', 'tvstand', 'stereo', 'window', 'dish_soap', 'couch', 'food_kiwi', 'wall', 'oven', 'bench', 'cup', 'coffe_maker', 'wine', 'tray', 'cards', 'dishrack', 'dresser', 'clothes_dress', 'maindoor', 'spoon', 'napkin', 'dishwasher', 'vase', 'colander']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the dishwasher. Walk to the dishwasher. Open the dishwasher. Find plate no.1. Grab plate no.1. Put plate no.1 into the dishwasher. Find plate no.2. Grab plate no.2. Put plate no.2 into the dishwasher. Find the spoon. Grab the spoon. Put the spoon into the dishwasher. Find the dish soap. Put the dish soap into the dishwasher. Close the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'blender', 'beer', 'mop', 'chair', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'ceiling', 'keyboard', 'alcohol', 'ceilinglamp', 'faucet', 'window', 'powersocket', 'toy', 'food_cheese', 'check', 'tray', 'console', 'mouse', 'pencil', 'kitchen_counter', 'hairbrush', 'hair', 'food_steak', 'bathtub', 'bathroom_cabinet', 'vase']\nTask:Answer door\nresponse:\n{\n    \"thoughts\": Wait until doorbell rings, look to see who it is outside and open the door if you feel so inclined.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <man>', '[FIND] <man>', '[GREET] <man>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'dog', 'freezer', 'walllamp', 'dough', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'mousepad', 'food_peanut_butter', 'clothes_pants', 'dining_room', 'wall', 'novel', 'coffe_maker', 'coffee_filter', 'water', 'shoes', 'controller', 'dishwasher', 'clothes_shirt', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Go to the kitchen and make coffee.", "actions": ["[RUN] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[PUTOBJBACK] <coffee>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'spectacles', 'filing_cabinet', 'food_salt', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'character', 'shower', 'stereo', 'microphone', 'wall', 'toy', 'cutting_board', 'cup', 'novel', 'food_food', 'check', 'food_onion', 'coffee_pot', 'console', 'shaving_cream', 'towel_rack', 'shoes', 'cd', 'hairbrush', 'maindoor', 'soap', 'coffee']\nTask:Read\nresponse:\n{\n    \"thoughts\": Read a book in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'chair', 'freezer', 'ceilingfan', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookmark', 'ceilinglamp', 'stereo', 'balanceball', 'microphone', 'wall', 'powersocket', 'box', 'cup', 'food_cheese', 'mouthwash', 'television', 'coffee_filter', 'tray', 'doorjamb', 'kitchen_counter', 'hairdryer', 'home_office', 'table', 'vase']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Grab groceries and walk to the fridge. Open fridge and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tea', 'toilet', 'jelly', 'band-aids', 'dog', 'walllamp', 'bathroom_counter', 'floor', 'bedroom', 'keyboard', 'mop_bucket', 'mat', 'ceilinglamp', 'shower', 'mousepad', 'couch', 'food_peanut_butter', 'dining_room', 'wall', 'pasta', 'powersocket', 'nightstand', 'razor', 'food_donut', 'tray', 'chessboard', 'controller', 'laundry_detergent', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk into the bathroom. I walk over to the toilet. I sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'bookshelf', 'keyboard', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'photoframe', 'check', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "you go into your office, switch on the light, find chair at desk, sit in the chair, switch on the computer and work.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <desk>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'crayon', 'chair', 'walllamp', 'bathroom', 'ceiling', 'bedroom', 'keyboard', 'bookshelf', 'clothes_jacket', 'toaster', 'deck_of_cards', 'faucet', 'mousepad', 'microphone', 'wall', 'tape', 'novel', 'food_food', 'check', 'food_donut', 'tray', 'towel_rack', 'dresser', 'pajamas', 'clothes_dress', 'hairdryer', 'bathroom_cabinet', 'home_office', 'table']\nTask:Grab things\nresponse:\n{\n    \"thoughts\": The robot will need some kind of hands or grabbing device to transport objects.,\n    \"actions\": ['[WALK] <novel>', '[TURNTO] <novel>', '[GRAB] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'chair', 'cat', 'drawing', 'floor', 'tea_bag', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'stovefan', 'window', 'couch', 'food_orange', 'wall', 'food_food', 'television', 'oven_mitts', 'food_onion', 'curtain', 'after_shave', 'comb', 'napkin', 'home_office', 'table', 'dishwasher', 'bathroom_cabinet', 'laser_pointer', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room,find the sofa,sat down in the sofa,take some relaxation.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'tea_bag', 'ceilinglamp', 'tvstand', 'shower', 'window', 'homework', 'couch', 'dustpan', 'food_orange', 'dining_room', 'wall', 'nightstand', 'photoframe', 'clothes_gloves', 'coffe_maker', 'check', 'television', 'mouse', 'pencil', 'table', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room, turn to the sofa. Sit on the sofa and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'freezer', 'orchid', 'phone', 'clothes_jacket', 'table_cloth', 'mat', 'plate', 'shower', 'clothes_socks', 'food_jam', 'bowl', 'broom', 'nightstand', 'oven', 'cup', 'food_food', 'television', 'after_shave', 'curtain', 'doorjamb', 'mouse', 'dresser', 'clothes_dress', 'maindoor', 'door', 'napkin', 'instrument_guitar', 'home_office', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room, sit down by the cat, put out hand, stroke the cat from head to tail", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'chair', 'dog', 'toothbrush', 'drawing', 'bills', 'ceiling', 'folder', 'desk', 'bedroom', 'remote_control', 'button', 'clothes_jacket', 'stovefan', 'mat', 'faucet', 'dining_room', 'broom', 'nightstand', 'wall', 'cutting_board', 'novel', 'mouse', 'kitchen_counter', 'cd', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I pick up the remote. I press button on remote.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['freezer', 'walllamp', 'ground_coffee', 'knifeblock', 'cpuscreen', 'longboard', 'floor', 'phone', 'bedroom', 'toaster', 'faucet', 'pillow', 'wall', 'bed', 'novel', 'water', 'tray', 'centerpiece', 'cards', 'kitchen_counter', 'shoes', 'dresser', 'cupboard', 'conditioner', 'sink', 'bag', 'napkin', 'table', 'dishwasher', 'board_game']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, grab a glass, walk to sink, turn on tap and pour water into glass, drink from glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'orchid', 'knifeblock', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'television', 'curtain', 'shaving_cream', 'dresser', 'home_office', 'table', 'bathroom_cabinet', 'shoe_rack']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch favorite shows, netflix, hulu,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'jelly', 'walllamp', 'trashcan', 'hanger', 'longboard', 'floor', 'desk', 'mail', 'toaster', 'plate', 'shower', 'balanceball', 'couch', 'wall', 'oven', 'box', 'cup', 'bed', 'coffe_maker', 'tray', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher', 'vase', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to Kitchen, find dishwasher. Find vessels to be washed. Open dishwasher, put vessels in dish washer, close dish washer and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <plate>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'microwave', 'dog', 'chair', 'cat', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'phone', 'character', 'keyboard', 'clothes_jacket', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'closetdrawer', 'oven', 'food_cheese', 'novel', 'toothbrush_holder', 'television', 'food_donut', 'console', 'controller', 'clothes_scarf', 'light', 'colander', 'board_game', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I walk to my cat, and I pet him,\n    \"actions\": ['[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'ceiling', 'toaster', 'ceilinglamp', 'mat', 'shower', 'faucet', 'tvstand', 'deck_of_cards', 'picture', 'clothes_pants', 'food_orange', 'food_kiwi', 'oven', 'bench', 'bed', 'food_food', 'television', 'wine', 'pencil', 'dresser', 'clothes_dress', 'bathroom_cabinet', 'light', 'dishwasher', 'vase', 'table', 'coffee']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I find a radio. I turn on the radio. I listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'toilet', 'blow_dryer', 'bathroom', 'trashcan', 'orchid', 'dvd_player', 'keyboard', 'mail', 'computer', 'food_bacon', 'stovefan', 'ceilinglamp', 'picture', 'window', 'couch', 'clothes_pants', 'oven', 'photoframe', 'coffe_maker', 'food_food', 'television', 'food_onion', 'coffee_pot', 'sink', 'bathroom_cabinet', 'light', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": grab the vesseles and wash,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dining_room>', '[WASH] <bowl>', '[WALK] <detergent>', '[FIND] <detergent>', '[RINSE] <detergent>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_pizza', 'filing_cabinet', 'scissors', 'chair', 'ceilingfan', 'dough', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'coffee_table', 'television', 'coin', 'wooden_spoon', 'curtain', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <coffee_table>", "[PUTBACK] <remote_control> <coffee_table>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'mop', 'tea', 'freezer', 'walllamp', 'floor', 'character', 'stovefan', 'toaster', 'mat', 'tvstand', 'stereo', 'window', 'electric_shaver', 'food_noodles', 'couch', 'soap', 'pillow', 'wall', 'bench', 'toilet_paper', 'food_food', 'doorjamb', 'diary', 'dresser', 'hairbrush', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to bathroom. find closet. Uset It. Clean it. Use Toilet paper,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[CLOSE] <dresser>', '[FIND] <toilet_paper>', '[WASH] <dresser>', '[FIND] <toilet_paper>', '[TOUCH] <toilet_paper>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'character', 'toaster', 'mat', 'laptop', 'faucet', 'tvstand', 'couch', 'dining_room', 'wall', 'powersocket', 'tape', 'cup', 'photoframe', 'coffe_maker', 'shaving_cream', 'towel_rack', 'cupboard', 'maindoor', 'sink', 'bag', 'bathroom_cabinet', 'light', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Going to the sink and scrubbing hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'detergent', 'trashcan', 'dough', 'clothes_underwear', 'folder', 'stamp', 'bookshelf', 'clothes_jacket', 'table_cloth', 'mousepad', 'couch', 'food_jam', 'clothes_hat', 'pillow', 'nightstand', 'bench', 'oven', 'cutting_board', 'bed', 'food_food', 'check', 'doorjamb', 'console', 'cupboard', 'home_office', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": i go find the phone. i pick up the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'jelly', 'detergent', 'cat', 'bathroom_counter', 'keyboard', 'food_bacon', 'piano_bench', 'faucet', 'window', 'couch', 'candle', 'pillow', 'nightstand', 'tape', 'food_food', 'coffe_maker', 'video_game_controller', 'ice', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cd', 'cupboard', 'hair', 'door', 'bathroom_cabinet', 'table', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the cat. Pet the cat with your hand.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_pizza', 'juice', 'crayon', 'blow_dryer', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'mat', 'tvstand', 'faucet', 'wall', 'powersocket', 'bench', 'tape', 'photoframe', 'clothes_gloves', 'food_donut', 'after_shave', 'doorjamb', 'dishrack', 'door', 'sink', 'bathroom_cabinet', 'soap']\nTask:Clean sink\nresponse:\n{\n    \"thoughts\": Take a premoistened cleaning wipe and wipe down top of sink and faucet.  Next clean the sink bowl.  Let air dry.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[WIPE] <faucet>', '[WIPE] <sink>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['trashcan', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'folder', 'keyboard', 'remote_control', 'alcohol', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'couch', 'pillow', 'wall', 'nightstand', 'cup', 'television', 'curtain', 'food_steak', 'napkin', 'table', 'clothes_shirt', 'creditcard']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Take TV remote to change TV Channel. Press channel/program +/- button on TV remote.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[TOUCH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'freezer', 'dough', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'bookmark', 'stovefan', 'mat', 'window', 'pillow', 'wall', 'nightstand', 'powersocket', 'bed', 'novel', 'coffe_maker', 'curtain', 'doorjamb', 'door', 'sink', 'light', 'laser_pointer']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I go to the bedroom and turn on the light. I lie in bed. I grab my book from the nightstand, and look for the bookmark to check where I last left off. I read for a while. When I'm done, I put the bookmark on the page I stop at and place the book back on the nightstand.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[READ] <novel>', '[PUTBACK] <bookmark> <novel>', '[FIND] <nightstand>', '[PUTBACK] <novel> <nightstand>']\n}\n", "input": "\nScene_Objects:['food_bread', 'freezer', 'walllamp', 'trashcan', 'food_carrot', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bedroom', 'character', 'ceilinglamp', 'tvstand', 'faucet', 'dustpan', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'photoframe', 'bed', 'food_food', 'television', 'food_donut', 'food_rice', 'doorjamb', 'shaving_cream', 'table', 'vase']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "walk to kitchen and seprate the groceries and arrange in fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'chair', 'drawing', 'floor', 'ceiling', 'bedroom', 'alcohol', 'toaster', 'stovefan', 'mat', 'shower', 'faucet', 'deck_of_cards', 'shoe-shine_kit', 'pillow', 'wall', 'oven', 'cup', 'coffe_maker', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'home_office', 'table', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Turn on the faucet.  Place glass under tap and fill with water.  Turn off faucet.  Drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['sponge', 'water_glass', 'pot', 'glue', 'walllamp', 'trashcan', 'knifeblock', 'ceiling', 'clothes_jacket', 'stovefan', 'toaster', 'mat', 'faucet', 'microphone', 'wall', 'cup', 'bed', 'food_food', 'check', 'water', 'food_donut', 'curtain', 'kitchen_counter', 'chessboard', 'cupboard', 'laundry_detergent', 'sink', 'spoon', 'bathroom_cabinet', 'dirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'toilet', 'chair', 'bathroom', 'drawing', 'orchid', 'bathroom_counter', 'phone', 'character', 'bedroom', 'toaster', 'mat', 'stereo', 'faucet', 'floor_lamp', 'shower', 'couch', 'dining_room', 'pasta', 'oven', 'powersocket', 'love_seat', 'newspaper', 'photoframe', 'coffe_maker', 'tray', 'dresser', 'bathtub', 'soap']\nTask:Read\nresponse:\n{\n    \"thoughts\": Sit on an easy chair.  Turn on music.  Take a book and read.  Turn on lamp at night.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>', '[FIND] <newspaper>', '[GRAB] <newspaper>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <newspaper>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'ceilingfan', 'drawing', 'needle', 'bills', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'candle', 'wall', 'nightstand', 'photoframe', 'mouthwash', 'razor', 'television', 'food_rice', 'curtain', 'towel_rack', 'headset', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go into the living room. Walk over to the sofa. Lie down on the sofa. Turn on and watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'dog', 'band-aids', 'walllamp', 'ground_coffee', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'phone', 'food_cake', 'computer', 'toaster', 'mat', 'faucet', 'window', 'couch', 'candle', 'pillow', 'wall', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'table', 'light', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room and stand in front of the sofa. Then I turn and sit on the sofa and lean my back against the pillow and my feet on the floor.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'beer', 'microwave', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'button', 'toaster', 'dustpan', 'dining_room', 'bowl', 'wall', 'coffe_maker', 'toothbrush_holder', 'doorjamb', 'towel_rack', 'kitchen_counter', 'chessboard', 'vacuum_cleaner', 'headset', 'controller', 'clothes_scarf', 'coffee_cup', 'dishwasher', 'vase', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to the kitchen at the coffee maker, grab coffee, pour coffee into coffee maker, put coffee into coffee maker, push coffee maker's button, wait to complete, grab coffee cup", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffee>", "[GRAB] <coffee>", "[OPEN] <coffe_maker>", "[POUR] <coffee> <coffe_maker>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'mat', 'window', 'clothes_socks', 'couch', 'pillow', 'towel', 'broom', 'wall', 'bench', 'oven', 'toy', 'bed', 'coffe_maker', 'television', 'wooden_spoon', 'curtain', 'rag', 'doorjamb', 'food_onion', 'towel_rack', 'mouse', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to living room. Find couch. Sit on couch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'freezer', 'bathroom', 'detergent', 'food_butter', 'longboard', 'bookshelf', 'stovefan', 'shower', 'pillow', 'towel', 'dining_room', 'nightstand', 'powersocket', 'toy', 'basket_for_clothes', 'newspaper', 'light_bulb', 'photoframe', 'food_food', 'coin', 'doorjamb', 'headset', 'shoes', 'hairbrush', 'cupboard', 'napkin', 'home_office']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk into my room and turn the radio on so that I can listen to music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'orchid', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'foundation', 'homework', 'closetdrawer', 'couch', 'pillow', 'broom', 'nightstand', 'wall', 'oven', 'light_bulb', 'television', 'curtain', 'dresser', 'bathroom_cabinet', 'table', 'dishwasher', 'creditcard']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'freezer', 'cat', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'iron', 'balanceball', 'mousepad', 'dustpan', 'pillow', 'broom', 'wall', 'oven', 'coffee_table', 'novel', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'mouse', 'pencil', 'light', 'board_game']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I walk to the living room and pick up a book from the coffee table. I find my favorite chair.  I sit in the armchair and open the book.  I read.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[WALK] <coffee_table>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'food_pizza', 'filing_cabinet', 'microwave', 'chair', 'walllamp', 'orchid', 'knifeblock', 'bathroom_counter', 'ceiling', 'phone', 'stamp', 'keyboard', 'thread', 'computer', 'stovefan', 'toaster', 'plate', 'shower', 'mousepad', 'form', 'wall', 'food_cheese', 'bed', 'console', 'headset', 'dresser', 'spoon', 'bathroom_cabinet', 'bag']\nTask:Walk to room\nresponse:\n{\n    \"thoughts\": Leave the living room and enter the hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'laptop', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'tape', 'cup', 'photoframe', 'food_food', 'mouse', 'vacuum_cleaner', 'chessboard', 'napkin', 'table', 'soap']\nTask:Work\nresponse:\n", "output": {"thoughts": "first ill go to my home office adn ill switch on my laptop and with the use of mouse and keyboard il run my computer", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <mouse>", "[TOUCH] <mouse>", "[RUN] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'bookmark', 'toaster', 'mat', 'tvstand', 'faucet', 'shower', 'laptop', 'mousepad', 'wall', 'oven', 'powersocket', 'towel_rack', 'kitchen_counter', 'headset', 'door', 'sink', 'bathtub', 'home_office', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I go to the bathroom. I locate the toilet. I walk to the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[WALK] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'food_salt', 'freezer', 'walllamp', 'trashcan', 'ceilingfan', 'cpuscreen', 'clothes_underwear', 'food_egg', 'phone', 'character', 'bedroom', 'envelope', 'clothes_jacket', 'mat', 'tvstand', 'couch', 'clothes_pants', 'newspaper', 'television', 'console', 'dishrack', 'cupboard', 'laundry_detergent', 'door', 'maindoor', 'bathroom_cabinet', 'instrument_guitar', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk into laundry room. Open washing machine door. Put dirty clothes into washing machine. Measure laundry detergent and place into washing machine. Turn dial to appropriate wash setting. Power on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[CLOSE] <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[OPEN] <washing_machine>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'beer', 'toilet', 'band-aids', 'nail_polish', 'ceilingfan', 'food_carrot', 'toothbrush', 'knife', 'shampoo', 'ceiling', 'bedroom', 'stamp', 'tea_bag', 'mop_bucket', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'homework', 'food_noodles', 'dining_room', 'bed', 'ice', 'mouse', 'spoon', 'home_office', 'bathroom_cabinet', 'vase', 'colander']\nTask:Shut front door\nresponse:\n{\n    \"thoughts\": Close the storm door. Lock the storm door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['food_salt', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'folder', 'faucet', 'couch', 'food_peanut_butter', 'wall', 'cup', 'photoframe', 'bed', 'check', 'oven_mitts', 'after_shave', 'coffee_pot', 'mouse', 'kitchen_counter', 'cd', 'conditioner', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'light', 'soap', 'tooth_paste', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "You walk to the bathroom and turn to the sink. Grab the soap and wash your hands, put the soap back.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[PUTOBJBACK] <soap>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_ice_cream', 'freezer', 'walllamp', 'milk', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'alcohol', 'computer', 'stovefan', 'ceilinglamp', 'faucet', 'mousepad', 'wall', 'oven', 'powersocket', 'cup', 'ice', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink', 'light']\nTask:Get something to drink\nresponse:\n{\n    \"thoughts\": Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[CLOSE] <freezer>', '[POUR] <milk> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['sponge', 'juice', 'chair', 'jelly', 'nail_polish', 'cat', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'iron', 'clothes_pants', 'food_kiwi', 'wall', 'photoframe', 'food_food', 'water', 'wooden_spoon', 'curtain', 'doorjamb', 'sink', 'bathroom_cabinet', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Wash the teeth", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <water>", "[TURNTO] <water>", "[POINTAT] <water>", "[FIND] <teeth>", "[WASH] <teeth>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_salt', 'chair', 'wallshelf', 'character', 'keyboard', 'food_cake', 'mop_bucket', 'stovefan', 'toaster', 'iron', 'shower', 'window', 'clothes_socks', 'closetdrawer', 'dining_room', 'broom', 'box', 'coffe_maker', 'television', 'tray', 'centerpiece', 'doorjamb', 'kettle', 'mouse', 'headset', 'cupboard', 'bag', 'table', 'light']\nTask:Shut front door\nresponse:\n{\n    \"thoughts\": Close the storm door. Lock the storm door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'ground_coffee', 'walllamp', 'food_butter', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'thread', 'stovefan', 'toaster', 'tvstand', 'faucet', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'coffe_maker', 'food_food', 'oven_mitts', 'coffee_filter', 'water', 'wine', 'wooden_spoon', 'controller', 'bathroom_cabinet', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'water_glass', 'toilet', 'jelly', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'stovefan', 'mat', 'shower', 'food_kiwi', 'wall', 'nightstand', 'oven', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'water', 'tray', 'doorjamb', 'coffee_pot', 'kitchen_counter', 'cupboard', 'door', 'food_steak', 'spoon', 'light', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_pizza', 'mop', 'bathroom', 'toothbrush', 'drawing', 'knifeblock', 'knife', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'mop_bucket', 'brush', 'toaster', 'faucet', 'stereo', 'couch', 'towel', 'wall', 'powersocket', 'cup', 'photoframe', 'bed', 'mouse', 'dresser', 'sink', 'bathroom_cabinet', 'soap', 'tooth_paste']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Upon entering the restroom, walk to the sink. Take a toothbrush and apply enough toothpaste so that 1/4 of the brush is covered (this is the appropriate amount). Proceed to \"brush\" your teeth using an up and down motion with the brush in your mouth. Brush for 2 minutes covering all areas of the teeth. Rinse with water and clean toothbrush with water too.", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[SCRUB] <tooth_paste>", "[FIND] <toothbrush>", "[WASH] <toothbrush>", "[FIND] <sink>", "[RINSE] <sink>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'stamp', 'bedroom', 'stovefan', 'toaster', 'plate', 'faucet', 'shower', 'dish_soap', 'food_jam', 'towel', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'ice', 'tray', 'curtain', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['jelly', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'brush', 'stovefan', 'faucet', 'iron', 'couch', 'towel', 'wall', 'powersocket', 'cutting_board', 'photoframe', 'bed', 'television', 'lighter', 'console', 'dresser', 'pajamas', 'door', 'sink', 'bathroom_cabinet', 'table', 'clothes_skirt', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go into the bathroom, get my hands wet, add soap, suds up my hands, rinse the soap off, turn the faucet off, dry my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[PUTOBJBACK] <soap>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'chair', 'freezer', 'trashcan', 'knife', 'longboard', 'floor', 'shampoo', 'computer', 'button', 'tvstand', 'stereo', 'window', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'nightstand', 'bed', 'novel', 'check', 'wine', 'curtain', 'cupboard', 'hair', 'sink', 'bathroom_cabinet', 'soap', 'creditcard']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I would go lie down in my bed and open the book and start reading.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'toilet', 'chair', 'walllamp', 'ceilingfan', 'toothbrush', 'drawing', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'keyboard', 'computer', 'bookmark', 'ceilinglamp', 'shower', 'faucet', 'window', 'mousepad', 'oil', 'couch', 'wall', 'nightstand', 'photoframe', 'novel', 'video_game_controller', 'television', 'mouse', 'table']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Find book, grab book, fin bookmark, sit down somewhere comfortable, open book, read book, close book after reading.", "actions": ["[WALK] <desk>", "[FIND] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[FIND] <couch>", "[SIT] <couch>", "[GRAB] <novel>", "[READ] <novel>", "[PUTBACK] <bookmark> <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'bathroom', 'detergent', 'food_butter', 'drawing', 'wallshelf', 'floor', 'desk', 'bookshelf', 'plate', 'faucet', 'picture', 'iron', 'window', 'dish_soap', 'food_jam', 'wall', 'nightstand', 'toy', 'cutting_board', 'food_food', 'television', 'curtain', 'pajamas', 'conditioner', 'light', 'dishwasher', 'tooth_paste', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to the kitchen.  Find the dishwasher.  Open the dishwasher.  Find the plates.  Load the plates into the dishwasher.  Find the soap.  Load the soap into the dishwasher.  Close the dishwasher.  Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'mop', 'freezer', 'walllamp', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'keyboard', 'remote_control', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'dining_room', 'wall', 'wall_clock', 'photoframe', 'food_food', 'television', 'curtain', 'console', 'pencil', 'table', 'light', 'clothes_skirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk into the living room. I sit on the couch and grab the remote. Then I turn on the TV and change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'dry_pasta', 'food_ice_cream', 'toilet', 'freezer', 'knifeblock', 'wallshelf', 'folder', 'desk', 'stamp', 'bedroom', 'computer', 'toaster', 'mat', 'floor_lamp', 'dustpan', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'check', 'curtain', 'console', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['detergent', 'food_butter', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'remote_control', 'thread', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'faucet', 'mousepad', 'couch', 'dining_room', 'wall', 'bed', 'food_food', 'television', 'food_donut', 'curtain', 'lighter', 'doorjamb', 'cards', 'sink', 'table', 'scrabble']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "turn tv on, watch TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'ground_coffee', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'toaster', 'plate', 'faucet', 'foundation', 'mousepad', 'food_jam', 'food_kiwi', 'wall', 'pasta', 'powersocket', 'food_food', 'coffe_maker', 'rag', 'shaving_cream', 'kitchen_counter', 'dishrack', 'dresser', 'cupboard', 'sink', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Take a dish and put it in the dishwasher to be washed,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'chair', 'freezer', 'trashcan', 'knife', 'longboard', 'floor', 'shampoo', 'computer', 'button', 'tvstand', 'stereo', 'window', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'nightstand', 'bed', 'novel', 'check', 'wine', 'curtain', 'cupboard', 'hair', 'sink', 'bathroom_cabinet', 'soap', 'creditcard']\nTask:Read book\nresponse:\n", "output": {"thoughts": "I would go lie down in my bed and open the book and start reading.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'crayon', 'chair', 'walllamp', 'bathroom', 'ceiling', 'bedroom', 'keyboard', 'bookshelf', 'clothes_jacket', 'toaster', 'deck_of_cards', 'faucet', 'mousepad', 'microphone', 'wall', 'tape', 'novel', 'food_food', 'check', 'food_donut', 'tray', 'towel_rack', 'dresser', 'pajamas', 'clothes_dress', 'hairdryer', 'bathroom_cabinet', 'home_office', 'table']\nTask:Grab things\nresponse:\n{\n    \"thoughts\": The robot will need some kind of hands or grabbing device to transport objects.,\n    \"actions\": ['[WALK] <novel>', '[TURNTO] <novel>', '[GRAB] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'bed', 'television', 'curtain', 'pencil', 'dresser', 'clothes_dress', 'conditioner', 'table', 'board_game', 'coffee']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Change the channel on the TV in the living room", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'walllamp', 'drawing', 'orchid', 'food_egg', 'floor', 'ceiling', 'phone', 'character', 'remote_control', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'shower', 'couch', 'clothes_hat', 'wall', 'powersocket', 'basket_for_clothes', 'check', 'mouthwash', 'television', 'food_onion', 'doorjamb', 'controller', 'door', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control,\n    \"actions\": ['[WALK] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['scissors', 'toilet', 'chair', 'walllamp', 'ceilingfan', 'bathroom_counter', 'floor', 'phone', 'character', 'mail', 'computer', 'toaster', 'foundation', 'clothes_socks', 'food_peanut_butter', 'dining_room', 'wall', 'oven', 'powersocket', 'tape', 'cup', 'coffe_maker', 'video_game_controller', 'television', 'tray', 'curtain', 'cd', 'maindoor', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'toilet', 'freezer', 'walllamp', 'drawing', 'knifeblock', 'floor', 'ceiling', 'food_chicken', 'food_cake', 'stovefan', 'mat', 'toaster', 'shower', 'food_noodles', 'oil', 'nightstand', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'light', 'dishwasher']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'scissors', 'food_salt', 'toilet', 'chair', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'mail', 'toaster', 'homework', 'mousepad', 'clothes_pants', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'food_food', 'coffee_pot', 'mouse', 'home_office', 'table', 'dishwasher', 'vase', 'tooth_paste']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom. Pull down pants. Use the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[PULL] <clothes_pants>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'food_carrot', 'drawing', 'orchid', 'cpuscreen', 'kitchen_cabinet', 'floor', 'ceiling', 'computer', 'mat', 'plate', 'ceilinglamp', 'stereo', 'couch', 'candle', 'food_orange', 'pillow', 'wall', 'oven', 'food_vegetable', 'novel', 'food_food', 'razor', 'television', 'console', 'kitchen_counter', 'cupboard', 'controller', 'table']\nTask:Fix snack\nresponse:\n{\n    \"thoughts\": Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTBACK] <food_carrot> <plate>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[PUTBACK] <food_vegetable> <plate>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['scissors', 'mop', 'dog', 'cat', 'knifeblock', 'longboard', 'folder', 'keyboard', 'food_cake', 'alcohol', 'table_cloth', 'mat', 'window', 'pillow', 'bench', 'toy', 'clothes_gloves', 'novel', 'coffe_maker', 'coin', 'towel_rack', 'vacuum_cleaner', 'dresser', 'cupboard', 'maindoor', 'door', 'table', 'dishwasher', 'laser_pointer', 'creditcard']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I pick up my phone. I touch it to navigate to the web browser. I can swipe and tap the screen to visit different websites. Then I read or watch the content within those websites.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>", "[TOUCH] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'freezer', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'shower', 'stereo', 'mousepad', 'wall', 'powersocket', 'food_food', 'food_donut', 'tray', 'after_shave', 'doorjamb', 'cards', 'mouse', 'shoes', 'dresser', 'food_steak', 'door', 'bathroom_cabinet', 'light', 'clothes_skirt']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Go kitchen and make something for eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_ice_cream', 'measuring_cup', 'knife', 'orchid', 'cpuscreen', 'bathroom_counter', 'ceiling', 'phone', 'alcohol', 'computer', 'table_cloth', 'mat', 'deck_of_cards', 'stereo', 'window', 'pillow', 'powersocket', 'cup', 'newspaper', 'check', 'ice', 'after_shave', 'lighter', 'diary', 'cd', 'cupboard', 'maindoor', 'colander', 'board_game']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find radio, turn it on", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'walllamp', 'bathroom', 'detergent', 'knifeblock', 'bathroom_counter', 'purse', 'alcohol', 'bookmark', 'stovefan', 'toaster', 'stereo', 'food_jam', 'dining_room', 'powersocket', 'food_food', 'mouthwash', 'tray', 'after_shave', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office', 'laser_pointer']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the hallway.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dining_room>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'beer', 'cat', 'trashcan', 'milk', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'dvd_player', 'character', 'desk', 'bookshelf', 'food_cake', 'computer', 'brush', 'ceilinglamp', 'plate', 'deck_of_cards', 'wall', 'powersocket', 'toy', 'newspaper', 'bed', 'coffe_maker', 'curtain', 'pencil', 'hair', 'door', 'light']\nTask:Give milk to cat\nresponse:\n", "output": {"thoughts": "Grab milk and pour it into the plate so that the cat can drink", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[POINTAT] <cat>", "[PULL] <cat>", "[FIND] <milk>", "[GRAB] <milk>", "[FIND] <plate>", "[POUR] <milk> <plate>", "[TOUCH] <cat>", "[SCRUB] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'blender', 'beer', 'freezer', 'band-aids', 'walllamp', 'orchid', 'food_egg', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'food_bacon', 'button', 'ceilinglamp', 'mousepad', 'food_peanut_butter', 'wall', 'powersocket', 'food_food', 'doorjamb', 'mouse', 'door', 'sink', 'napkin', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_pizza', 'tablelamp', 'microwave', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'envelope', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'deck_of_cards', 'window', 'couch', 'wall', 'nightstand', 'novel', 'television', 'curtain', 'hairdryer', 'sink', 'home_office', 'table', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the television. Switch on the television. Find the remote control. Grab the remote control. Point at television. Find the button. Push button.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'chair', 'cat', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'food_bacon', 'stovefan', 'ceilinglamp', 'iron', 'dustpan', 'pillow', 'wall', 'toy', 'box', 'centerpiece', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cd', 'dresser', 'cupboard', 'light', 'table']\nTask:Vacuum\nresponse:\n{\n    \"thoughts\": Plug in the vacuum. Turn it on. Push it back and forth across the floor until all the debris is gone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'dog', 'chair', 'bathroom', 'drawing', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'purse', 'mat', 'ceilinglamp', 'shower', 'candle', 'wall', 'bed', 'clothes_gloves', 'food_food', 'water', 'food_rice', 'wooden_spoon', 'curtain', 'console', 'dishrack', 'pajamas', 'light', 'clothes_skirt', 'soap', 'dirt']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "go to a bathroom, open the shower, let water wash your body", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <soap>", "[SCRUB] <soap>", "[FIND] <water>", "[RINSE] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'scissors', 'chair', 'hanger', 'phone', 'keyboard', 'bookshelf', 'plate', 'faucet', 'window', 'mousepad', 'food_noodles', 'oil', 'food_peanut_butter', 'clothes_pants', 'dining_room', 'bench', 'oven', 'cutting_board', 'toy', 'basket_for_clothes', 'cup', 'photoframe', 'keys', 'food_food', 'laundry_detergent', 'bathroom_cabinet', 'instrument_violin', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'beer', 'measuring_cup', 'toilet', 'drawing', 'knife', 'hanger', 'floor', 'desk', 'tea_bag', 'keyboard', 'piano_bench', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'foundation', 'clothes_pants', 'towel', 'wall', 'nightstand', 'novel', 'television', 'tray', 'curtain', 'bag', 'table', 'vase', 'soap']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "BEFORE EATING", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <towel>", "[WIPE] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'crayon', 'ceilingfan', 'drawing', 'needle', 'floor', 'ceiling', 'shampoo', 'keyboard', 'remote_control', 'thread', 'food_bacon', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'photoframe', 'bed', 'television', 'coffee_filter', 'curtain', 'mouse', 'table', 'light', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go into the living room. I find the remote. I turn on the tv. I sit on the couch and watch tv.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'crayon', 'mop', 'toilet', 'toothbrush', 'drawing', 'bathroom_counter', 'floor', 'desk', 'computer', 'ceilinglamp', 'deck_of_cards', 'faucet', 'mousepad', 'microphone', 'towel', 'wall', 'oven', 'tape', 'photoframe', 'novel', 'toothbrush_holder', 'towel_rack', 'cupboard', 'sink', 'bathroom_cabinet', 'home_office', 'table', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "getting ready for bed", "actions": ["[WALK] <bathroom_counter>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <toothbrush> <bathroom_counter>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>", "[GRAB] <toothbrush>", "[SCRUB] <toothbrush>", "[RINSE] <toothbrush>", "[FIND] <toothbrush_holder>", "[PUTBACK] <toothbrush> <toothbrush_holder>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'food_bread', 'tablelamp', 'crayon', 'tea', 'toilet', 'walllamp', 'knife', 'floor', 'ceiling', 'toaster', 'mat', 'plate', 'faucet', 'dish_soap', 'couch', 'wall', 'oven', 'novel', 'coffe_maker', 'drinking_glass', 'tray', 'doorjamb', 'kettle', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'dishwasher', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen and turn to the dishwasher, put the knife, glass, fork and plate into the dishwasher. Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[TURNTO] <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[OPEN] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['food_bread', 'tablelamp', 'chair', 'nail_polish', 'detergent', 'ceilingfan', 'knifeblock', 'orchid', 'cpuscreen', 'hanger', 'character', 'keyboard', 'computer', 'homework', 'window', 'towel', 'cutting_board', 'bed', 'photoframe', 'curtain', 'rag', 'towel_rack', 'mouse', 'kitchen_counter', 'headset', 'dresser', 'controller', 'maindoor', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room, find lamp, switch on.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'food_salt', 'measuring_cup', 'toilet', 'jelly', 'drawing', 'floor', 'phone', 'ceiling', 'character', 'bookshelf', 'alcohol', 'mat', 'shower', 'foundation', 'electric_shaver', 'mousepad', 'wall', 'toy', 'basket_for_clothes', 'photoframe', 'wine', 'dresser', 'laundry_detergent', 'food_steak', 'door', 'napkin', 'bathroom_cabinet', 'dishwasher']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Pickup toys from around the room. Organize toys back into their appropriate spaces on the shelves and into the toy boxes.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTBACK] <toy> <dresser>', '[CLOSE] <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'walllamp', 'knife', 'food_egg', 'floor', 'food_cake', 'toaster', 'plate', 'tvstand', 'window', 'dish_soap', 'wall', 'nightstand', 'oven', 'cutting_board', 'coffe_maker', 'check', 'razor', 'television', 'food_donut', 'tray', 'centerpiece', 'doorjamb', 'dishrack', 'food_steak', 'spoon', 'light', 'dishwasher', 'colander']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "put forks, spoons, plates, knife in dishwasher, add dish soap, turn diswaser on", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[FIND] <dishrack>", "[PUTBACK] <fork> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishrack>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishrack>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'toothbrush', 'drawing', 'orchid', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'bedroom', 'stovefan', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'mousepad', 'pillow', 'wall', 'pasta', 'oven', 'coffe_maker', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'sauce_pan', 'cupboard']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Cook some pasta on the stove,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pasta>', '[GRAB] <pasta>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <sauce_pan>', '[POUR] <water> <sauce_pan>', '[FIND] <oven>', '[SWITCHON] <oven>', '[PUTBACK] <pasta> <sauce_pan>', '[FIND] <sauce_pan>', '[GRAB] <sauce_pan>', '[PUTBACK] <sauce_pan> <oven>']\n}\n", "input": "\nScene_Objects:['tea', 'chair', 'freezer', 'cpuscreen', 'food_egg', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'stovefan', 'mousepad', 'wall', 'keys', 'check', 'oven_mitts', 'doorjamb', 'mouse', 'pencil', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'napkin', 'home_office', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "turn on the light for room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[SWITCHON] <light>", "[FIND] <chair>", "[SIT] <chair>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'walllamp', 'bathroom', 'drawing', 'ceiling', 'dvd_player', 'character', 'bookshelf', 'keyboard', 'brush', 'ceilinglamp', 'toaster', 'mat', 'window', 'nightstand', 'bench', 'photoframe', 'food_food', 'after_shave', 'diary', 'kettle', 'shoes', 'chessboard', 'dresser', 'door', 'bathroom_cabinet', 'colander']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Using soft brush, carefully dust both Indonesian wedding doll statues.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <brush>', '[GRAB] <brush>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['food_salt', 'bathroom', 'detergent', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_peanut_butter', 'pillow', 'towel', 'dining_room', 'wall', 'powersocket', 'toilet_paper', 'television', 'curtain', 'pencil', 'headset', 'cd', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Lift the remote and change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'ceilingfan', 'knifeblock', 'orchid', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'computer', 'couch', 'broom', 'wall', 'coffe_maker', 'video_game_controller', 'wine', 'doorjamb', 'towel_rack', 'cards', 'pencil', 'kitchen_counter', 'mouse', 'dresser', 'cupboard', 'door', 'light', 'board_game']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk over to light switch. Turn on lights.,\n    \"actions\": ['[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'toilet', 'chair', 'cat', 'trashcan', 'ceilingfan', 'orchid', 'cpuscreen', 'longboard', 'desk', 'remote_control', 'button', 'food_bacon', 'foundation', 'couch', 'dining_room', 'oven', 'bench', 'basket_for_clothes', 'cup', 'bed', 'check', 'toothbrush_holder', 'doorjamb', 'coffee_pot', 'vacuum_cleaner', 'controller', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk into living room. Find television remote control. Pick up remote control. Press button.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'beer', 'freezer', 'walllamp', 'bathroom', 'bathroom_counter', 'character', 'bedroom', 'desk', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'clothes_pants', 'towel', 'bench', 'basket_for_clothes', 'bed', 'coin', 'food_donut', 'curtain', 'towel_rack', 'mouse', 'cd', 'comb', 'laundry_detergent', 'bathtub', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'microwave', 'freezer', 'ceilingfan', 'knifeblock', 'drawing', 'orchid', 'floor', 'ceiling', 'dvd_player', 'character', 'bookshelf', 'clothes_jacket', 'ceilinglamp', 'tvstand', 'deck_of_cards', 'mousepad', 'couch', 'food_jam', 'pillow', 'food_kiwi', 'wall', 'powersocket', 'bed', 'food_food', 'ice', 'mouse', 'door', 'dishwasher', 'scrabble']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk Into Kitchen. Open Refrigerator. Grab Food. Walk to Microwave. Insert food into microwave. Cook until warm.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[CLOSE] <freezer>", "[WALK] <microwave>", "[SWITCHON] <microwave>", "[SWITCHOFF] <microwave>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'bathroom', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'purse', 'stovefan', 'mat', 'toaster', 'faucet', 'shower', 'plate', 'closetdrawer', 'towel', 'wall', 'wall_clock', 'oven', 'toy', 'hands_both', 'coin', 'tray', 'mouse', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom. Turn on the sink. Rinse hands. Apply soap and scrub your hands together. Rinse hands again. Turn off faucet. Dry hands on a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <hands_both>', '[RINSE] <hands_both>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['freezer', 'bathroom', 'ceilingfan', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'stovefan', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'cutting_board', 'bed', 'novel', 'mouthwash', 'television', 'curtain', 'towel_rack', 'home_office', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I come home and go to the sofa for turn off the tv and relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'band-aids', 'dough', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'dvd_player', 'desk', 'bedroom', 'keyboard', 'computer', 'clothes_jacket', 'ceilinglamp', 'mat', 'mousepad', 'wall', 'powersocket', 'cup', 'bed', 'photoframe', 'newspaper', 'water', 'food_donut', 'doorjamb', 'mouse', 'door', 'light', 'table', 'shoe_rack']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <freezer>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <cup>', '[POUR] <water> <cup>', '[FIND] <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'mop', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'envelope', 'stovefan', 'faucet', 'couch', 'form', 'wall', 'bench', 'photoframe', 'coffe_maker', 'television', 'food_donut', 'wine', 'tray', 'after_shave', 'mouse', 'paper_towel', 'door', 'sink', 'bathroom_cabinet', 'dishwasher', 'clothes_skirt', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk to the bathroom and find the sink. I grab the soap and rub some on my hands. I turn the faucet on and put my hands in the water. I rinse my hands and then turn off the faucet. Then I dry my hands on a paper towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[RINSE] <hands_both>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'bills', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'dining_room', 'wall', 'nightstand', 'powersocket', 'oven', 'photoframe', 'food_food', 'television', 'doorjamb', 'towel_rack', 'mouse', 'bathroom_cabinet', 'table', 'home_office']\nTask:Work\nresponse:\n{\n    \"thoughts\": Sit at a desk.  Turn on computer.  Sign onto mTurk.  Perform tasks.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'cd_player', 'freezer', 'detergent', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'phone', 'character', 'button', 'toaster', 'ceilinglamp', 'stovefan', 'window', 'food_noodles', 'couch', 'wall', 'wall_clock', 'bench', 'powersocket', 'television', 'towel_rack', 'dresser', 'sink', 'table', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room, walk to cd player, turn on cd player, press play button, walk to sofa, sit in sofa and listen to music", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <button>", "[PUSH] <button>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_pizza', 'pot', 'microwave', 'bathroom', 'bills', 'knifeblock', 'cpuscreen', 'drawing', 'longboard', 'floor', 'phone', 'folder', 'bedroom', 'keyboard', 'mail', 'food_bacon', 'piano_bench', 'button', 'table_cloth', 'shower', 'couch', 'dining_room', 'wall', 'powersocket', 'cutting_board', 'water', 'tray', 'pencil', 'door']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Drinking a glass of water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'chair', 'walllamp', 'bathroom', 'trashcan', 'ceilingfan', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'laptop', 'balanceball', 'mousepad', 'pillow', 'wall', 'cup', 'bed', 'television', 'razor', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'spoon', 'light', 'vase']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "push any ligth button to turn it on, for example in the living room", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'knifeblock', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'desk', 'food_cake', 'stovefan', 'toaster', 'deck_of_cards', 'faucet', 'candle', 'pillow', 'dining_room', 'oven', 'photoframe', 'check', 'food_food', 'television', 'bathroom_cabinet', 'table', 'dishwasher', 'soap', 'board_game', 'clothes_shirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk to living room. Find cat. Walk to cat. Pull cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[PULL] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'jelly', 'freezer', 'chair', 'drawing', 'clothes_underwear', 'bathroom_counter', 'floor', 'button', 'stovefan', 'ceilinglamp', 'brush', 'deck_of_cards', 'window', 'oil', 'couch', 'clothes_hat', 'dining_room', 'broom', 'wall', 'mouthwash', 'television', 'kitchen_counter', 'dishrack', 'sink', 'bathroom_cabinet', 'table', 'home_office', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room ,find sofa ,sit on sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['nail_polish', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'faucet', 'couch', 'dining_room', 'wall', 'nightstand', 'bench', 'wall_clock', 'photoframe', 'coffe_maker', 'food_food', 'lighter', 'doorjamb', 'towel_rack', 'diary', 'shoes', 'comb', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom. turn on the water in the sink. apply soap to hands. rub hands together. rinse hands with water. turn off the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['spectacles', 'cpuscreen', 'bathroom_counter', 'longboard', 'phone', 'desk', 'bookshelf', 'computer', 'stovefan', 'tvstand', 'shower', 'balanceball', 'window', 'foundation', 'clothes_socks', 'bowl', 'newspaper', 'bed', 'check', 'food_food', 'television', 'razor', 'food_rice', 'cards', 'cupboard', 'maindoor', 'napkin', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TOUCH] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_pizza', 'filing_cabinet', 'scissors', 'chair', 'ceilingfan', 'dough', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'wall', 'powersocket', 'coffee_table', 'television', 'coin', 'wooden_spoon', 'curtain', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'food_ice_cream', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'longboard', 'floor', 'bookshelf', 'computer', 'faucet', 'balanceball', 'window', 'iron', 'food_peanut_butter', 'food_jam', 'wall', 'nightstand', 'bench', 'powersocket', 'basket_for_clothes', 'novel', 'check', 'coffee_filter', 'shaving_cream', 'pencil', 'maindoor', 'sink', 'hairdryer']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Enter the Home office, then walk to the phone, pickup the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'detergent', 'food_butter', 'toothbrush', 'drawing', 'floor', 'ceiling', 'toaster', 'mat', 'plate', 'faucet', 'shower', 'dish_soap', 'wall', 'nightstand', 'cup', 'coffe_maker', 'coffee_filter', 'drinking_glass', 'tray', 'wooden_spoon', 'kitchen_counter', 'cupboard', 'sink', 'coffee_cup', 'dishwasher', 'shoe_rack']\nTask:Clean dishes\nresponse:\n{\n    \"thoughts\": Wash dishes in the sink then place them in the dishwasher. Put detergent in the dishwasher and turn it on when full.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <drinking_glass>', '[WASH] <drinking_glass>', '[FIND] <coffee_cup>', '[WASH] <coffee_cup>', '[FIND] <plate>', '[WASH] <plate>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <plate>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <dishwasher>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <detergent>', '[GRAB] <detergent>', '[PUTBACK] <detergent> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'food_ice_cream', 'beer', 'microwave', 'freezer', 'walllamp', 'floor', 'ceiling', 'phone', 'dvd_player', 'stovefan', 'ceilinglamp', 'toaster', 'mousepad', 'wall', 'wall_clock', 'cup', 'bed', 'food_food', 'coin', 'food_rice', 'water', 'pencil', 'shoes', 'controller', 'napkin', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Feeling tired after hard work I opened the fridge and grabbed cool water and drank from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <cup>", "[GRAB] <cup>", "[CLOSE] <freezer>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <cup>", "[POUR] <water> <cup>", "[FIND] <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'freezer', 'bathroom', 'trashcan', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'mail', 'stovefan', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'pillow', 'wall_clock', 'child', 'check', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'dresser', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Say hi,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>']\n}\n", "input": "\nScene_Objects:['pot', 'blender', 'crayon', 'measuring_cup', 'freezer', 'nail_polish', 'cat', 'floor', 'dvd_player', 'character', 'bedroom', 'tea_bag', 'piano_bench', 'toaster', 'foundation', 'window', 'couch', 'photoframe', 'check', 'toilet_paper', 'curtain', 'centerpiece', 'shaving_cream', 'mouse', 'chessboard', 'headset', 'sink', 'napkin', 'spoon', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "pick up  the book open the book and read", "actions": ["[WALK] <bedroom>", "[WALK] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'toilet', 'walllamp', 'nail_polish', 'dough', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'cards', 'dishrack', 'door', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": first i will go to living room and then take the remote and swtitch on television after that il change the channel,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TOUCH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['fork', 'filing_cabinet', 'blow_dryer', 'trashcan', 'drawing', 'knife', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'toaster', 'faucet', 'picture', 'towel', 'dining_room', 'wall', 'nightstand', 'powersocket', 'photoframe', 'novel', 'food_food', 'controller', 'door', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'soap', 'board_game']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. Walk to the sink. Turn on tap. Get soap. Wash and rinse hands. Turn off tap. Dry hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[LOOKAT] <sink>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[RINSE] <soap>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'detergent', 'orchid', 'wallshelf', 'bedroom', 'tea_bag', 'bookshelf', 'ceilinglamp', 'laptop', 'faucet', 'window', 'couch', 'broom', 'wall', 'oven', 'cutting_board', 'toy', 'cup', 'standingmirror', 'razor', 'television', 'tray', 'after_shave', 'towel_rack', 'mouse', 'kitchen_counter', 'controller', 'bathtub', 'vase', 'soap']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk to living room, walk to the phone reach for the phone and pick up the phone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[TOUCH] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'freezer', 'blow_dryer', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'shoe-shine_kit', 'couch', 'microphone', 'food_jam', 'wall', 'coffe_maker', 'check', 'television', 'curtain', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'colander', 'tooth_paste']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the television. Walk to the television. Switch on the television. Find the sofa. Walk to the sofa. Sit on the sofa. Find the remote control. Grab the remote control. Look at the television.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[WALK] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'chair', 'bathroom', 'detergent', 'trashcan', 'toothbrush', 'knifeblock', 'bathroom_counter', 'character', 'keyboard', 'computer', 'food_bacon', 'tvstand', 'nightstand', 'oven', 'bench', 'powersocket', 'cup', 'bed', 'ice', 'diary', 'cards', 'dresser', 'hairbrush', 'home_office', 'dishwasher', 'scrabble']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": go to radio. run on radio.,\n    \"actions\": ['[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_ice_cream', 'cd_player', 'orchid', 'bedroom', 'purse', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'stereo', 'tvstand', 'balanceball', 'clothes_socks', 'couch', 'food_jam', 'powersocket', 'oven', 'cutting_board', 'photoframe', 'novel', 'check', 'television', 'wine', 'food_onion', 'pencil', 'vacuum_cleaner', 'door', 'sink', 'instrument_violin']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk in the living room, then I go the the stereo and the cd player and turn them on.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_bread', 'juice', 'toilet', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'food_peanut_butter', 'food_jam', 'bowl', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'food_cheese', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'food_steak', 'light', 'dishwasher', 'table']\nTask:Grab some juice\nresponse:\n{\n    \"thoughts\": I go to the fridge, and grab some juice out of it. I then get a glass, and pour the juice into the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <juice>', '[GRAB] <juice>', '[POUR] <juice> <cup>', '[PUTOBJBACK] <juice>', '[CLOSE] <freezer>', '[FIND] <table>', '[PUTBACK] <cup> <table>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'toilet', 'chair', 'freezer', 'floor', 'ceiling', 'thread', 'table_cloth', 'ceilinglamp', 'faucet', 'food_peanut_butter', 'dining_room', 'wall', 'nightstand', 'powersocket', 'tape', 'photoframe', 'bed', 'food_food', 'cards', 'kitchen_counter', 'dishrack', 'dresser', 'hair', 'pajamas', 'door', 'home_office', 'dishwasher', 'vase']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "go to the kitchen and put the groceries in the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'freezer', 'trashcan', 'drawing', 'bathroom_counter', 'ceiling', 'phone', 'thread', 'computer', 'stovefan', 'faucet', 'shower', 'food_jam', 'dining_room', 'wall', 'powersocket', 'tape', 'coffe_maker', 'food_food', 'mirror', 'doorjamb', 'mouse', 'paper_towel', 'dresser', 'cupboard', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'home_office', 'dirt']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": Spray Windex or glass and mirror cleaner onto mirror. Use paper towel to wipe off cleaner until mirror is dry and has no streaks.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[WIPE] <mirror>', '[FIND] <trashcan>', '[PUTBACK] <paper_towel> <trashcan>']\n}\n", "input": "\nScene_Objects:['toilet', 'detergent', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'keyboard', 'thread', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'mousepad', 'couch', 'wall', 'photoframe', 'bed', 'toilet_paper', 'food_onion', 'towel_rack', 'mouse', 'kitchen_counter', 'cd', 'hairbrush', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "walk to bathroom , run to toilet", "actions": ["[WALK] <bathroom>", "[RUN] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'detergent', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mail', 'button', 'clothes_jacket', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'wall_clock', 'powersocket', 'keys', 'television', 'ice', 'doorjamb', 'mouse', 'pencil', 'headset', 'dresser', 'sink', 'light', 'table']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Switch on computer, Use keybord and mouse to open e-mail provider, type requested e-mail, click send, close e-mail provider, turn off computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[FIND] <keyboard>', '[FIND] <mail>', '[OPEN] <mail>', '[TYPE] <keyboard>', '[FIND] <button>', '[PUSH] <button>', '[CLOSE] <mail>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'ceilingfan', 'drawing', 'needle', 'floor', 'ceiling', 'folder', 'character', 'keyboard', 'remote_control', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'couch', 'dining_room', 'wall', 'pasta', 'light_bulb', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'shoes', 'table', 'shoe_rack']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk into the Living room.  Find the television remote.  Use it to turn on the television.  Point the remote at the television. Press the channel button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'filing_cabinet', 'blender', 'toilet', 'drawing', 'needle', 'food_egg', 'clothes_underwear', 'bathroom_counter', 'phone', 'keyboard', 'computer', 'table_cloth', 'closetdrawer', 'clothes_hat', 'nightstand', 'oven', 'basket_for_clothes', 'photoframe', 'check', 'television', 'coin', 'food_donut', 'diary', 'towel_rack', 'headset', 'comb', 'hairdryer', 'napkin']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Drink water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['fork', 'sheets', 'chair', 'dog', 'walllamp', 'orchid', 'cpuscreen', 'longboard', 'phone', 'character', 'desk', 'table_cloth', 'clothes_hat', 'towel', 'wall_clock', 'oven', 'bed', 'light_bulb', 'food_food', 'drinking_glass', 'water', 'curtain', 'after_shave', 'console', 'headset', 'cupboard', 'maindoor', 'spoon', 'bathroom_cabinet', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. Take a glass out of the cabinet. Then I walk to the sink and turn the water on and fill the glass. Then I take a drink.", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <drinking_glass>", "[POUR] <water> <drinking_glass>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'jelly', 'chair', 'phone', 'ceiling', 'desk', 'bedroom', 'table_cloth', 'bookmark', 'stovefan', 'mat', 'tvstand', 'stereo', 'shoe-shine_kit', 'faucet', 'toaster', 'mousepad', 'powersocket', 'cutting_board', 'shelf', 'food_cheese', 'television', 'wooden_spoon', 'after_shave', 'food_onion', 'doorjamb', 'shoes', 'dishrack', 'vase', 'electrical_outlet']\nTask:Turn on radio\nresponse:\n{\n    \"thoughts\": Go to the radio.  Switch on the power button.  Tune the station.  Set the volume.  Listen to the music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[FIND] <electrical_outlet>', '[TURNTO] <electrical_outlet>', '[PLUGIN] <stereo>', '[SWITCHON] <electrical_outlet>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_salt', 'toilet', 'chair', 'food_butter', 'ceilingfan', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'character', 'desk', 'bedroom', 'bookshelf', 'keyboard', 'food_cake', 'computer', 'shoe-shine_kit', 'faucet', 'mousepad', 'oil', 'wall', 'nightstand', 'bed', 'mouse', 'kitchen_counter', 'vacuum_cleaner', 'cupboard', 'sink', 'bathroom_cabinet']\nTask:Read book\nresponse:\n", "output": {"thoughts": "go to living room,find the text book on the self ,sat on the chair ,open and read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'dining_room', 'wall', 'powersocket', 'oven', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Push power button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'beer', 'freezer', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'food_cake', 'button', 'ceilinglamp', 'mat', 'pillow', 'wall', 'cup', 'basket_for_clothes', 'bed', 'check', 'coffe_maker', 'food_donut', 'tray', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'food_steak', 'table', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen ,find fridge ,open fridge ,find water  glass , drink from water glass", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'desk', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'powersocket', 'clothes_gloves', 'television', 'razor', 'curtain', 'doorjamb', 'console', 'mouse', 'door', 'bag', 'table', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I walk in living room. I find TV remote. I grab TV remote. I sit on sofa. I switch on TV. I switch channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['juice', 'food_ice_cream', 'blow_dryer', 'detergent', 'ceilingfan', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'shoe-shine_kit', 'couch', 'food_orange', 'wall', 'tape', 'television', 'curtain', 'kitchen_counter', 'conditioner', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "For relaxing walk to living room. Find the sofa to relax. Arrange the sofa to relax. Sit on the sofa to relaxing.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <conditioner>", "[FIND] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'walllamp', 'ceilingfan', 'drawing', 'hanger', 'floor', 'bedroom', 'computer', 'ceilinglamp', 'deck_of_cards', 'iron', 'window', 'electric_shaver', 'couch', 'food_orange', 'wall', 'wall_clock', 'bed', 'novel', 'check', 'television', 'toothbrush_holder', 'wooden_spoon', 'after_shave', 'towel_rack', 'dresser', 'conditioner', 'maindoor', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'drawing', 'clothes_underwear', 'bathroom_counter', 'ceiling', 'purse', 'clothes_jacket', 'table_cloth', 'stovefan', 'shoe-shine_kit', 'faucet', 'shower', 'window', 'clothes_socks', 'clothes_pants', 'microphone', 'wall', 'nightstand', 'pasta', 'photoframe', 'razor', 'wine', 'hairbrush', 'clothes_dress', 'hairdryer', 'sink', 'bathroom_cabinet', 'dishwasher', 'vase', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open cabinet and collect clothes and put them into washing machine and wash clothes", "actions": ["[WALK] <bathroom>", "[WALK] <bathroom_cabinet>", "[FIND] <bathroom_cabinet>", "[OPEN] <bathroom_cabinet>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'orchid', 'knifeblock', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'television', 'curtain', 'shaving_cream', 'dresser', 'home_office', 'table', 'bathroom_cabinet', 'shoe_rack']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch favorite shows, netflix, hulu,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTBACK] <remote_control> <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'envelope', 'clothes_jacket', 'toaster', 'faucet', 'wall', 'oven', 'coffe_maker', 'television', 'drinking_glass', 'tray', 'curtain', 'lighter', 'coffee_pot', 'kitchen_counter', 'paper_towel', 'dishrack', 'comb', 'dresser', 'cupboard', 'sink', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "wipe loose dirt, place in the dishwasher, turn on the dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>", "[WIPE] <drinking_glass>", "[RINSE] <drinking_glass>", "[PUTBACK] <drinking_glass> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'toilet', 'detergent', 'dough', 'bills', 'wallshelf', 'clothes_underwear', 'ceiling', 'keyboard', 'table_cloth', 'faucet', 'electric_shaver', 'mousepad', 'couch', 'dustpan', 'form', 'dining_room', 'newspaper', 'photoframe', 'food_food', 'coffe_maker', 'check', 'wooden_spoon', 'tray', 'rag', 'instrument_guitar', 'soap', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sponge', 'chair', 'freezer', 'walllamp', 'food_carrot', 'drawing', 'knife', 'hanger', 'floor', 'phone', 'ceiling', 'keyboard', 'computer', 'brush', 'deck_of_cards', 'food_jam', 'bowl', 'dining_room', 'wall', 'bench', 'razor', 'doorjamb', 'mouse', 'dishrack', 'dresser', 'cupboard', 'door', 'sink', 'table', 'light']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Pick up phone", "actions": ["[WALK] <table>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'ground_coffee', 'food_butter', 'food_carrot', 'bills', 'floor', 'ceiling', 'bedroom', 'computer', 'table_cloth', 'tvstand', 'shower', 'faucet', 'wall', 'oven', 'newspaper', 'photoframe', 'bed', 'coffee_filter', 'curtain', 'kitchen_counter', 'cupboard', 'controller', 'dishwasher', 'clothes_shirt']\nTask:Make popcorn\nresponse:\n{\n    \"thoughts\": Lay popcorn bag flat inside microwave and press popcorn button to start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <microwave>', '[TURNTO] <microwave>', '[OPEN] <microwave>', '[PUTIN] <food_food> <microwave>', '[CLOSE] <microwave>', '[SWITCHON] <microwave>']\n}\n", "input": "\nScene_Objects:['freezer', 'ground_coffee', 'nail_polish', 'bathroom', 'food_butter', 'bills', 'phone', 'bedroom', 'food_cake', 'envelope', 'bookmark', 'ceilinglamp', 'tvstand', 'stereo', 'shower', 'balanceball', 'dining_room', 'food_kiwi', 'powersocket', 'toy', 'cup', 'photoframe', 'food_food', 'television', 'tray', 'dishrack', 'maindoor', 'home_office', 'dishwasher', 'vase']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk into living room, walk to stereo, turn on stereo, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'microwave', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'bathroom_counter', 'phone', 'desk', 'bookshelf', 'toaster', 'ceilinglamp', 'shower', 'faucet', 'couch', 'dining_room', 'nightstand', 'wall_clock', 'oven', 'wall', 'bed', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'bathtub', 'bathroom_cabinet', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom, look at the water, wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <water>', '[FIND] <water>', '[TURNTO] <water>', '[LOOKAT] <water>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'toilet', 'bathroom', 'trashcan', 'ceilingfan', 'toothbrush', 'orchid', 'floor', 'phone', 'bookshelf', 'computer', 'toaster', 'shoe-shine_kit', 'mousepad', 'pillow', 'dining_room', 'box', 'basket_for_clothes', 'keys', 'razor', 'toothbrush_holder', 'coin', 'wine', 'doorjamb', 'cards', 'door', 'napkin', 'home_office', 'soap', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "today morning i went to bathroom and took my toothbrush and paste from tooth brush holder and brushed and cleaned my teeth well", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[PUTBACK] <tooth_paste> <toothbrush_holder>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'bathroom', 'bills', 'orchid', 'cpuscreen', 'floor', 'phone', 'character', 'bedroom', 'piano_bench', 'toaster', 'mat', 'ceilinglamp', 'window', 'foundation', 'electric_shaver', 'oil', 'closetdrawer', 'couch', 'broom', 'nightstand', 'wall', 'cup', 'bed', 'check', 'toothbrush_holder', 'sink', 'napkin', 'dishwasher']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <sheets>', '[PULL] <sheets>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'beer', 'microwave', 'chair', 'bathroom', 'trashcan', 'dough', 'knifeblock', 'floor', 'computer', 'stovefan', 'mat', 'balanceball', 'faucet', 'window', 'couch', 'candle', 'wall', 'toy', 'bed', 'photoframe', 'coffe_maker', 'video_game_controller', 'wooden_spoon', 'curtain', 'dresser', 'pajamas', 'hair', 'maindoor']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "go to bedroom, put pajamas on, lie in bed, sleep", "actions": ["[WALK] <bedroom>", "[WALK] <pajamas>", "[FIND] <pajamas>", "[GRAB] <pajamas>", "[PUTON] <pajamas>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'toaster', 'mat', 'stovefan', 'shower', 'faucet', 'window', 'couch', 'pillow', 'wall', 'oven', 'coffe_maker', 'television', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'food_steak', 'door', 'sink', 'table']\nTask:Make toast\nresponse:\n{\n    \"thoughts\": Place slices of bread in a toaster ,wait for the bread to pop up .Set aside.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[CLOSE] <freezer>', '[FIND] <toaster>', '[PLUGIN] <toaster>', '[PUTBACK] <food_bread> <toaster>', '[SWITCHON] <toaster>']\n}\n", "input": "\nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'nail_polish', 'ceilingfan', 'drawing', 'food_egg', 'floor', 'bedroom', 'keyboard', 'toaster', 'plate', 'stereo', 'shower', 'dish_soap', 'microphone', 'wall', 'bench', 'oven', 'box', 'check', 'video_game_controller', 'food_food', 'coffe_maker', 'tray', 'kettle', 'console', 'table', 'dishwasher', 'soap']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go the in kitchen then grab the plate on the table. I put the plate and dish soap in the dishwasher. Finally, I turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'dog', 'freezer', 'drawing', 'cpuscreen', 'phone', 'tea_bag', 'food_bacon', 'stovefan', 'mat', 'ceilinglamp', 'window', 'oil', 'food_orange', 'food_cheese', 'clothes_gloves', 'toilet_paper', 'food_food', 'television', 'coffee_filter', 'check', 'curtain', 'kettle', 'mouse', 'food_steak', 'spoon', 'instrument_guitar', 'light', 'dishwasher']\nTask:Walk to room\nresponse:\n{\n    \"thoughts\": Leave the living room and enter the hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'food_salt', 'ceilingfan', 'dough', 'food_carrot', 'orchid', 'desk', 'remote_control', 'mail', 'button', 'computer', 'bookmark', 'mop_bucket', 'ceilinglamp', 'tvstand', 'iron', 'shower', 'couch', 'broom', 'bench', 'tape', 'check', 'tray', 'curtain', 'mouse', 'pencil', 'cupboard', 'controller', 'scrabble']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pick up the remote. I press button on remote.", "actions": ["[WALK] <bedroom>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'nail_polish', 'food_carrot', 'drawing', 'cpuscreen', 'floor', 'phone', 'desk', 'keyboard', 'stovefan', 'mat', 'tvstand', 'homework', 'window', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'oven', 'bench', 'bed', 'novel', 'coffe_maker', 'toothbrush_holder', 'towel_rack', 'home_office', 'instrument_violin']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Go to the bedroom. Grab the book out of the nightstand. Lay down in bed. Open the book. Read the book.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <nightstand>', '[FIND] <nightstand>', '[OPEN] <nightstand>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <bed>', '[TURNTO] <bed>', '[SIT] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'blender', 'freezer', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'plate', 'iron', 'wall', 'nightstand', 'bed', 'food_food', 'check', 'drinking_glass', 'curtain', 'centerpiece', 'coffee_pot', 'cupboard', 'laundry_detergent', 'door', 'dishwasher', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Robot Drinks the water", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>", "[PUTOBJBACK] <drinking_glass>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'tablelamp', 'microwave', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_bacon', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'food_food', 'television', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": go to the kitchen and cook some thing to eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <oven>', '[OPEN] <oven>', '[FIND] <food_pizza>', '[GRAB] <food_pizza>', '[PUTIN] <food_pizza> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'food_bread', 'toilet', 'orchid', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'longboard', 'shampoo', 'desk', 'stamp', 'keyboard', 'toaster', 'mat', 'deck_of_cards', 'faucet', 'foundation', 'mousepad', 'form', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'cupboard', 'door', 'bathroom_cabinet', 'light']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Travel through the house and locate the bathroom, and proceed to the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'microwave', 'toilet', 'toothbrush', 'bills', 'orchid', 'character', 'bedroom', 'stamp', 'stovefan', 'laptop', 'ceilinglamp', 'shower', 'mat', 'faucet', 'towel', 'pillow', 'dining_room', 'oven', 'coffe_maker', 'television', 'toothbrush_holder', 'comb', 'cupboard', 'door', 'sink', 'table', 'soap', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Find toothbrush.  Find toothpaste.  Put toothpaste on toothbrush.  Brush teeth.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toothbrush_holder>', '[FIND] <toothbrush_holder>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[WALK] <tooth_paste>', '[FIND] <tooth_paste>', '[GRAB] <tooth_paste>', '[POUR] <tooth_paste> <toothbrush>', '[FIND] <teeth>', '[SCRUB] <teeth>']\n}\n", "input": "\nScene_Objects:['fork', 'food_bread', 'tablelamp', 'mop', 'chair', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'phone', 'desk', 'bookmark', 'plate', 'shower', 'tvstand', 'pillow', 'wall', 'bench', 'newspaper', 'check', 'food_food', 'doorjamb', 'console', 'mouse', 'chessboard', 'dresser', 'spoon', 'table', 'light', 'vase']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Place your plate on the placemat. The forks will be placed on the left side of the plate and the knife and spoon on the right and place the spoon to the right of the knife. If you want to set out food, place them above the plate.", "actions": ["[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[DROP] <plate>", "[FIND] <fork>", "[GRAB] <fork>", "[DROP] <fork>", "[FIND] <spoon>", "[GRAB] <spoon>", "[DROP] <spoon>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'walllamp', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'bookshelf', 'mail', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'oven', 'toy', 'mouthwash', 'television', 'wine', 'tray', 'curtain', 'cd', 'dresser', 'bathtub', 'sink', 'table', 'light']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room,find the sofa,sat down in the sofa,take some relaxation.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'microwave', 'chair', 'detergent', 'ceilingfan', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'clothes_jacket', 'mat', 'deck_of_cards', 'pillow', 'powersocket', 'photoframe', 'check', 'food_food', 'food_rice', 'wooden_spoon', 'doorjamb', 'shaving_cream', 'kettle', 'cd', 'shoes', 'controller', 'home_office', 'dishwasher', 'board_game', 'clothes_shirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find Cat in living room.  Walk to cat.  Pet cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'trashcan', 'drawing', 'orchid', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'computer', 'toaster', 'deck_of_cards', 'mousepad', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'powersocket', 'bench', 'television', 'doorjamb', 'cd', 'dresser', 'conditioner', 'door', 'home_office', 'light', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the phone. I pick up the phone. I hold it to my ear.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['sheets', 'chair', 'freezer', 'cat', 'cpuscreen', 'hanger', 'longboard', 'floor', 'ceiling', 'desk', 'mail', 'clothes_jacket', 'ceilinglamp', 'iron', 'picture', 'oil', 'food_noodles', 'clothes_pants', 'wall', 'novel', 'food_food', 'coffe_maker', 'television', 'tray', 'headset', 'maindoor', 'napkin', 'vase', 'tooth_paste', 'dirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Open fridge. Put groceries in fridge.", "actions": ["[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'toilet', 'freezer', 'cpuscreen', 'clothes_underwear', 'wallshelf', 'shampoo', 'character', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'clothes_pants', 'dining_room', 'bench', 'powersocket', 'photoframe', 'coffe_maker', 'mouse', 'paper_towel', 'dresser', 'door', 'bathtub', 'bathroom_cabinet', 'light', 'hands_both']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walking towards Toilet room, And dispose of human waste. Cleaning your hands with tissue paper or water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <paper_towel>', '[FIND] <paper_towel>', '[TURNTO] <paper_towel>', '[LOOKAT] <paper_towel>', '[GRAB] <paper_towel>', '[FIND] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['blow_dryer', 'trashcan', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookmark', 'toaster', 'stovefan', 'tvstand', 'food_jam', 'pillow', 'dining_room', 'bowl', 'wall', 'powersocket', 'clothes_gloves', 'novel', 'coffe_maker', 'coffee_filter', 'tray', 'after_shave', 'curtain', 'cd', 'chessboard', 'cupboard', 'sink', 'dishwasher', 'scrabble', 'coffee']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Enter kitchen, find coffee maker, find coffee filter, pick up coffee filter, put it in coffee maker, find coffee, pour coffee into filter, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'toilet', 'chair', 'detergent', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mail', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'novel', 'food_onion', 'doorjamb', 'mouse', 'napkin', 'light', 'clothes_skirt']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to the living room, where my computer is. I locate it, switch it on and type desired URL adress in the web browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'bedroom', 'envelope', 'stovefan', 'brush', 'faucet', 'mousepad', 'towel', 'pillow', 'wall', 'photoframe', 'coffe_maker', 'kitchen_counter', 'headset', 'instrument_violin', 'sink', 'maindoor', 'bathroom_cabinet', 'light', 'home_office', 'table', 'soap', 'board_game', 'hands_both', 'creditcard']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to bathroom, turn on faucet, wet hands, apply soap, rub hands, rinse hands, turn off faucet, dry hands", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'bathroom', 'ceilingfan', 'knife', 'drawing', 'needle', 'knifeblock', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'phone', 'piano_bench', 'faucet', 'window', 'teeth', 'clothes_pants', 'wall', 'photoframe', 'check', 'oven_mitts', 'tray', 'pencil', 'clothes_dress', 'sink', 'bathroom_cabinet', 'home_office', 'clothes_skirt', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[PUSH] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'microwave', 'jelly', 'detergent', 'food_carrot', 'toothbrush', 'needle', 'hanger', 'floor', 'phone', 'shampoo', 'character', 'tvstand', 'iron', 'balanceball', 'stereo', 'mousepad', 'food_jam', 'pillow', 'nightstand', 'toothbrush_holder', 'diary', 'console', 'shaving_cream', 'kitchen_counter', 'sink', 'light', 'dishwasher', 'vase']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the playback device you want to listen to. turn on. press play. enjoyu music.", "actions": ["[WALK] <bedroom>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'character', 'desk', 'computer', 'food_bacon', 'mat', 'tvstand', 'ceilinglamp', 'plate', 'window', 'shower', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'wine', 'curtain', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_bread', 'microwave', 'walllamp', 'ceilingfan', 'food_butter', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'food_orange', 'wall', 'bench', 'photoframe', 'keys', 'check', 'food_food', 'television', 'curtain', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Go to the living room, Switch on the television. Find the remote and press the change channel button, change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[TOUCH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'walllamp', 'nail_polish', 'trashcan', 'drawing', 'orchid', 'floor', 'desk', 'computer', 'food_bacon', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'bench', 'coffe_maker', 'television', 'curtain', 'mouse', 'bathtub', 'home_office', 'table', 'bathroom_cabinet']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'oil', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'coffe_maker', 'television', 'curtain', 'food_onion', 'controller', 'clothes_dress', 'table', 'clothes_skirt', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'shower', 'window', 'mousepad', 'dining_room', 'wall', 'check', 'television', 'wine', 'doorjamb', 'shaving_cream', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Complete surveys on amazon turk\nresponse:\n{\n    \"thoughts\": Sit down at computer, Turn computer on. Go online to Amazon mechanical turk, Browse through surveys, Select a survey.,\n    \"actions\": ['[WALK] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'toilet', 'freezer', 'nail_polish', 'detergent', 'longboard', 'floor', 'ceiling', 'bedroom', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'food_peanut_butter', 'form', 'wall', 'nightstand', 'oven', 'food_food', 'ice', 'after_shave', 'doorjamb', 'lighter', 'coffee_pot', 'mouse', 'spoon']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Put groceries into the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'clothes_jacket', 'table_cloth', 'mat', 'floor_lamp', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'light_bulb', 'novel', 'mouthwash', 'television', 'ice', 'curtain', 'after_shave', 'doorjamb', 'console', 'kettle', 'conditioner', 'bathtub', 'sink', 'table', 'vase']\nTask:Read book\nresponse:\n{\n    \"thoughts\": A person brings a book into the living room and sits down in a comfortable chair.  If it is dark in the room the person turns on the lamp. The book is opened to the next page to be read.  The person begins reading the book until they need to stop for some reason.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'mop', 'band-aids', 'walllamp', 'trashcan', 'ceilingfan', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'desk', 'computer', 'stovefan', 'toaster', 'deck_of_cards', 'shower', 'wall', 'pasta', 'oven', 'cup', 'water', 'tray', 'doorjamb', 'shaving_cream', 'kitchen_counter', 'sauce_pan', 'cupboard', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Cook some pasta on the stove", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pasta>", "[GRAB] <pasta>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <sauce_pan>", "[POUR] <water> <sauce_pan>", "[FIND] <oven>", "[SWITCHON] <oven>", "[PUTBACK] <pasta> <sauce_pan>", "[FIND] <sauce_pan>", "[GRAB] <sauce_pan>", "[PUTBACK] <sauce_pan> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'dough', 'wallshelf', 'floor', 'ceiling', 'bedroom', 'bookmark', 'stovefan', 'brush', 'deck_of_cards', 'faucet', 'window', 'closetdrawer', 'dining_room', 'wall', 'oven', 'cutting_board', 'cup', 'bed', 'coffe_maker', 'food_food', 'tray', 'shaving_cream', 'kitchen_counter', 'chessboard', 'clothes_dress', 'conditioner']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[TURNTO] <cup>', '[LOOKAT] <cup>', '[WALK] <cup>', '[FIND] <freezer>', '[TURNTO] <freezer>', '[LOOKAT] <freezer>', '[WALK] <freezer>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'detergent', 'food_butter', 'ceilingfan', 'drawing', 'orchid', 'clothes_underwear', 'floor', 'ceiling', 'dvd_player', 'character', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'nightstand', 'bench', 'coffe_maker', 'television', 'curtain', 'console', 'kitchen_counter', 'door', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the living room. Sit on the sofa. Grab the remote and turn on the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'crayon', 'walllamp', 'trashcan', 'dough', 'floor', 'phone', 'bedroom', 'tea_bag', 'toaster', 'mat', 'plate', 'ceilinglamp', 'dish_soap', 'wall', 'pasta', 'oven', 'cutting_board', 'food_cheese', 'newspaper', 'coffe_maker', 'check', 'food_food', 'tray', 'towel_rack', 'pencil', 'dresser', 'conditioner', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'laptop', 'shoe-shine_kit', 'balanceball', 'picture', 'ceilinglamp', 'mousepad', 'wall', 'nightstand', 'newspaper', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I need to finish some work on my computer in my home office to meet a deadline for tomorrow.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <desk>", "[OPEN] <desk>", "[PUTBACK] <keyboard> <desk>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <mousepad>", "[PUTBACK] <mouse> <mousepad>", "[PLUGIN] <keyboard>", "[PLUGIN] <mouse>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[TYPE] <keyboard>", "[GRAB] <mouse>", "[TOUCH] <mouse>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'dvd_player', 'character', 'bookshelf', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'laptop', 'couch', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'lighter', 'dresser', 'sink', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'juice', 'toilet', 'bills', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'character', 'desk', 'keyboard', 'bookshelf', 'table_cloth', 'brush', 'window', 'mousepad', 'clothes_hat', 'wall', 'nightstand', 'photoframe', 'razor', 'tray', 'mouse', 'cd', 'table', 'dishwasher', 'colander', 'scrabble', 'teeth']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "i go to the bathroom. I find the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'walllamp', 'ceilingfan', 'drawing', 'hanger', 'floor', 'bedroom', 'computer', 'ceilinglamp', 'deck_of_cards', 'iron', 'window', 'electric_shaver', 'couch', 'food_orange', 'wall', 'wall_clock', 'bed', 'novel', 'check', 'television', 'toothbrush_holder', 'wooden_spoon', 'after_shave', 'towel_rack', 'dresser', 'conditioner', 'maindoor', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['water_glass', 'food_bread', 'bathroom', 'trashcan', 'food_carrot', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'phone', 'folder', 'bedroom', 'desk', 'keyboard', 'toaster', 'bench', 'tray', 'doorjamb', 'diary', 'console', 'mouse', 'towel_rack', 'cd', 'dresser', 'cupboard', 'controller', 'conditioner', 'table', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, walk to table, grab water glass and drink water", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>", "[PUTBACK] <water_glass> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'tea', 'microwave', 'chair', 'walllamp', 'blow_dryer', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'plate', 'mousepad', 'food_noodles', 'couch', 'closetdrawer', 'food_jam', 'form', 'pillow', 'wall', 'nightstand', 'razor', 'television', 'shaving_cream', 'kitchen_counter', 'bathroom_cabinet', 'dishwasher', 'soap', 'teeth']\nTask:Text friends while sitting on couch\nresponse:\n{\n    \"thoughts\": Text while sitting,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <phone>', '[LOOKAT] <phone>', '[TOUCH] <phone>']\n}\n", "input": "\nScene_Objects:['blow_dryer', 'trashcan', 'drawing', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'clothes_jacket', 'brush', 'tvstand', 'faucet', 'balanceball', 'iron', 'wall', 'bench', 'photoframe', 'check', 'razor', 'television', 'food_rice', 'console', 'hairdryer', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'instrument_guitar', 'home_office', 'vase', 'teeth', 'shoe_rack']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom.  Find mouthwash or turn on water.  Put mouthwash or water in mouth.  Swish around for 30 seconds.  Spit mouthwash out.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <teeth>", "[RINSE] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'drawing', 'needle', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'couch', 'dustpan', 'dining_room', 'wall', 'nightstand', 'powersocket', 'food_food', 'toilet_paper', 'doorjamb', 'coffee_pot', 'mouse', 'kettle', 'dresser', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk into living room, walk to light, turn on light switch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'toilet', 'freezer', 'cat', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'oil', 'couch', 'clothes_hat', 'wall', 'basket_for_clothes', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'towel_rack', 'comb', 'dresser', 'maindoor', 'soap']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I put my groceries into the fridge.", "actions": ["[WALK] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'stovefan', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'nightstand', 'tape', 'newspaper', 'bed', 'television', 'doorjamb', 'mouse', 'sink', 'home_office', 'table', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office. Sit on chair. Switch on computer. type in keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'tea', 'freezer', 'walllamp', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'folder', 'stovefan', 'ceilinglamp', 'toaster', 'microphone', 'wall', 'bench', 'oven', 'cup', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'controller', 'table', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and open the fridge. From the fridge take some chicken and close the fridge. Walk to the table and put the chicken on the table. Plug in oven and put the chicken into the oven.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[CLOSE] <freezer>", "[WALK] <table>", "[PUTBACK] <food_chicken> <table>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <table>", "[FIND] <oven>", "[PLUGIN] <oven>", "[OPEN] <oven>", "[GRAB] <food_chicken>", "[PUTIN] <food_chicken> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'oil', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'coffe_maker', 'television', 'curtain', 'food_onion', 'controller', 'clothes_dress', 'table', 'clothes_skirt', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I get home from work and decide to watch tv. I go to the living room and turn on the tv. I sit on the couch and use the remote to switch the channels to find what I want to watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'bathroom', 'trashcan', 'knife', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'food_bacon', 'tvstand', 'faucet', 'oil', 'clothes_pants', 'wall', 'oven', 'photoframe', 'newspaper', 'after_shave', 'shaving_cream', 'sink', 'bathroom_cabinet', 'light', 'clothes_skirt', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Walk to the sink, turn on the faucet, put soap on hands, rub hands together under the water, rinse off hands, turn off faucet, dry hands", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[WALK] <sink>", "[FIND] <soap>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'blender', 'food_salt', 'crayon', 'tea', 'cat', 'milk', 'knifeblock', 'hanger', 'shampoo', 'bedroom', 'bookshelf', 'computer', 'brush', 'stovefan', 'toaster', 'foundation', 'window', 'mousepad', 'clothes_pants', 'clothes_hat', 'tape', 'bed', 'coffe_maker', 'food_onion', 'after_shave', 'dresser', 'light', 'dishwasher', 'laser_pointer']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Every time this animal is seen, you must approach it and show compassion for it by petting it in a gentle manner.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <bedroom>', '[WALK] <home_office>', '[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[FIND] <brush>', '[GRAB] <brush>']\n}\n", "input": "\nScene_Objects:['mop', 'toilet', 'ceilingfan', 'orchid', 'drawing', 'cpuscreen', 'knifeblock', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'food_cake', 'clothes_jacket', 'table_cloth', 'stovefan', 'mousepad', 'pillow', 'dining_room', 'wall', 'nightstand', 'cup', 'photoframe', 'clothes_gloves', 'light_bulb', 'toilet_paper', 'coin', 'ice', 'mouse', 'cupboard', 'vase']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bath room and go to the toilet", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'tablelamp', 'ground_coffee', 'walllamp', 'trashcan', 'knifeblock', 'longboard', 'floor', 'ceiling', 'character', 'bookmark', 'toaster', 'tvstand', 'wall', 'clothes_gloves', 'keys', 'coffe_maker', 'check', 'video_game_controller', 'coffee_filter', 'toilet_paper', 'food_onion', 'curtain', 'kitchen_counter', 'controller', 'hairdryer', 'dishwasher', 'vase', 'soap']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": You walk into the kitchen and turn to the coffee maker. You grab the coffee filter and put into the coffee machine, you also grab the ground coffee and put it into the coffee machine. You turn on the coffee machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[OPEN] <coffe_maker>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['fork', 'tablelamp', 'walllamp', 'bathroom', 'trashcan', 'food_carrot', 'clothes_underwear', 'floor', 'desk', 'toaster', 'plate', 'shower', 'tvstand', 'food_noodles', 'clothes_socks', 'mousepad', 'wall', 'oven', 'cup', 'newspaper', 'coffe_maker', 'toilet_paper', 'food_donut', 'water', 'tray', 'food_onion', 'lighter', 'maindoor', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "first i will take all the dishes to the dish washer and with the help of dishwasher and water i will clean all the dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water>", "[TURNTO] <water>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'ground_coffee', 'walllamp', 'knife', 'knifeblock', 'hanger', 'bathroom_counter', 'floor', 'desk', 'bedroom', 'button', 'stovefan', 'mat', 'shower', 'clothes_socks', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'bed', 'light_bulb', 'check', 'video_game_controller', 'mouthwash', 'food_onion', 'vacuum_cleaner', 'pajamas', 'cupboard', 'hair', 'instrument_guitar']\nTask:Go to sleep\nresponse:\n{\n    \"thoughts\": go to bedroom, put pajamas on, lie in bed, sleep,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <pajamas>', '[FIND] <pajamas>', '[GRAB] <pajamas>', '[PUTON] <pajamas>', '[FIND] <bed>', '[LIE] <bed>']\n}\n", "input": "\nScene_Objects:['food_salt', 'toilet', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'desk', 'bedroom', 'keyboard', 'tvstand', 'stereo', 'balanceball', 'picture', 'mousepad', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'photoframe', 'bed', 'toilet_paper', 'coffe_maker', 'cards', 'mouse', 'kitchen_counter', 'cd', 'vase']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[OPEN] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'microwave', 'walllamp', 'bathroom', 'detergent', 'knifeblock', 'knife', 'clothes_underwear', 'kitchen_cabinet', 'floor', 'ceiling', 'phone', 'toaster', 'stovefan', 'plate', 'faucet', 'tvstand', 'wall', 'nightstand', 'oven', 'coffe_maker', 'drinking_glass', 'tray', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'home_office', 'dishwasher']\nTask:Unload dishwasher\nresponse:\n{\n    \"thoughts\": Open the dishwasher door. Take out all of the dishes and silverware. Put the dishes and silverware in their drawers and cabinets. Close the dishwasher door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[PUTBACK] <drinking_glass> <cupboard>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[PUTIN] <fork> <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[PUTIN] <knife> <kitchen_cabinet>', '[FIND] <spoon>', '[GRAB] <spoon>', '[PUTIN] <spoon> <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTIN] <plate> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>']\n}\n", "input": "\nScene_Objects:['fork', 'chair', 'cloth_napkin', 'dog', 'nail_polish', 'dough', 'drawing', 'knife', 'hanger', 'bathroom_counter', 'floor', 'phone', 'computer', 'plate', 'bowl', 'food_kiwi', 'wall', 'bench', 'powersocket', 'cup', 'bed', 'video_game_controller', 'television', 'curtain', 'doorjamb', 'dresser', 'cupboard', 'door', 'table', 'light']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk to dining room, walk to dining table, arrange chairs, set napkin, set plate, set knife, set fork and glass on table.", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <chair>", "[PULL] <chair>", "[FIND] <cloth_napkin>", "[GRAB] <cloth_napkin>", "[PUTBACK] <cloth_napkin> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'food_salt', 'chair', 'walllamp', 'toothbrush', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'oil', 'closetdrawer', 'couch', 'food_jam', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'check', 'television', 'doorjamb', 'dresser', 'door']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to living room. Walk to and sit at the sofa. Grab remote control and switch on TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['toilet', 'dog', 'walllamp', 'bathroom', 'orchid', 'longboard', 'floor', 'phone', 'desk', 'bookshelf', 'keyboard', 'food_bacon', 'table_cloth', 'bookmark', 'piano_bench', 'shower', 'floor_lamp', 'homework', 'mousepad', 'food_kiwi', 'bed', 'photoframe', 'curtain', 'doorjamb', 'kitchen_counter', 'dishrack', 'dresser', 'napkin', 'table', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'juice', 'tea', 'toilet', 'freezer', 'walllamp', 'dough', 'drawing', 'food_egg', 'floor', 'dvd_player', 'character', 'alcohol', 'stovefan', 'mat', 'shower', 'closetdrawer', 'clothes_hat', 'food_jam', 'wall', 'powersocket', 'curtain', 'doorjamb', 'mouse', 'kitchen_counter', 'laundry_detergent', 'door', 'light', 'dishwasher', 'table']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[TURNTO] <toilet>', '[LOOKAT] <toilet>', '[WALK] <toilet>', '[TURNTO] <toilet>']\n}\n", "input": "\nScene_Objects:['fork', 'water_glass', 'pot', 'freezer', 'walllamp', 'bathroom', 'food_butter', 'food_carrot', 'milk', 'knifeblock', 'knife', 'floor', 'ceiling', 'mail', 'computer', 'stovefan', 'ceilinglamp', 'brush', 'faucet', 'wall', 'nightstand', 'cutting_board', 'cup', 'photoframe', 'water', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'soap']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I enter the kitchen and grab a waterglass from the cupboard. I open the fridge and grap some water. I pour the water into the cup and drink it.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[OPEN] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'microwave', 'toilet', 'walllamp', 'drawing', 'bills', 'wallshelf', 'floor', 'ceiling', 'character', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'television', 'tray', 'curtain', 'towel_rack', 'pencil', 'kitchen_counter', 'bathtub', 'spoon', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'tablelamp', 'microwave', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'bench', 'cup', 'doorjamb', 'mouse', 'address_book', 'hairbrush', 'hairdryer', 'light', 'creditcard']\nTask:Work\nresponse:\n", "output": {"thoughts": "Walk to home office. Walk to computer. Switch on computer. Open the address book.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <address_book>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'cat', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'mail', 'alcohol', 'mat', 'faucet', 'clothes_pants', 'pillow', 'wall', 'wall_clock', 'oven', 'pasta', 'bench', 'coffee_filter', 'ice', 'paper_towel', 'conditioner', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk to the bathroom and find the sink. I grab the soap and rub some on my hands. I turn the faucet on and put my hands in the water. I rinse my hands and then turn off the faucet. Then I dry my hands on a paper towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <hands_both>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'fork', 'filing_cabinet', 'tablelamp', 'blender', 'mop', 'bathroom', 'detergent', 'toothbrush', 'bills', 'hanger', 'floor', 'bedroom', 'envelope', 'toaster', 'stovefan', 'clothes_socks', 'mousepad', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'toilet_paper', 'towel_rack', 'clothes_dress', 'bag', 'home_office', 'light', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to Bathroom. Find basket for clothes. Lift clothes dress from basket. Find washing machine. Walk to washing machine. Put clothes in washing machine. Put soap in washing machine. Switch on washing machine.", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[FIND] <washing_machine>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'mop', 'microwave', 'toilet', 'chair', 'dog', 'walllamp', 'floor', 'desk', 'bookshelf', 'keyboard', 'mat', 'plate', 'shower', 'window', 'bowl', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'shelf', 'check', 'food_food', 'cards', 'towel_rack', 'kitchen_counter', 'maindoor', 'table', 'vase', 'dirt']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['blender', 'food_ice_cream', 'microwave', 'toilet', 'freezer', 'ceilingfan', 'knife', 'character', 'desk', 'bookshelf', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'couch', 'candle', 'clothes_pants', 'food_orange', 'wall', 'nightstand', 'bench', 'tape', 'check', 'curtain', 'doorjamb', 'dishrack', 'clothes_dress', 'food_steak', 'scrabble', 'coffee']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Pick up the book, open to current page, read the words.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'toilet', 'freezer', 'food_butter', 'knife', 'floor', 'table_cloth', 'toaster', 'plate', 'faucet', 'stereo', 'dish_soap', 'oil', 'wall', 'wall_clock', 'oven', 'powersocket', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'coffee_filter', 'wooden_spoon', 'tray', 'kitchen_counter', 'sink', 'dishwasher', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'food_ice_cream', 'measuring_cup', 'chair', 'freezer', 'knifeblock', 'hanger', 'floor', 'ceiling', 'phone', 'food_cake', 'computer', 'toaster', 'tvstand', 'food_noodles', 'couch', 'wall', 'powersocket', 'toy', 'tray', 'after_shave', 'doorjamb', 'chessboard', 'dresser', 'cupboard', 'hairbrush', 'door', 'napkin', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk into living room, walk to light, turn on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'scissors', 'crayon', 'toilet', 'jelly', 'walllamp', 'nail_polish', 'detergent', 'food_cake', 'alcohol', 'piano_bench', 'toaster', 'food_noodles', 'garbage_can', 'pillow', 'broom', 'oven', 'toy', 'food_vegetable', 'food_food', 'television', 'oven_mitts', 'doorjamb', 'pencil', 'headset', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'pantry']\nTask:Organize pantry\nresponse:\n{\n    \"thoughts\": Remove all perishable foods from the pantry, check dates and throw away any old food. Decide on an efficient ad neat arrangement for the food and replace it in the pantry accordingly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <pantry>', '[FIND] <food_food>', '[GRAB] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[FIND] <garbage_can>', '[PUTBACK] <food_food> <garbage_can>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <dry_pasta>', '[GRAB] <dry_pasta>', '[TURNTO] <dry_pasta>', '[LOOKAT] <dry_pasta>', '[PUTOBJBACK] <dry_pasta>', '[FIND] <food_food>', '[TURNTO] <food_food>', '[LOOKAT] <food_food>', '[PUSH] <food_food>', '[FIND] <food_noodles>', '[GRAB] <food_noodles>', '[TURNTO] <food_noodles>', '[LOOKAT] <food_noodles>', '[PUTOBJBACK] <food_noodles>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'ground_coffee', 'bathroom', 'trashcan', 'food_butter', 'knifeblock', 'needle', 'floor', 'ceiling', 'character', 'clothes_jacket', 'toaster', 'ceilinglamp', 'balanceball', 'mousepad', 'broom', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'coffee_filter', 'water', 'rag', 'coffee_pot', 'laundry_detergent', 'maindoor', 'napkin', 'table', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'toilet', 'freezer', 'walllamp', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'plate', 'faucet', 'drying_rack', 'dish_soap', 'couch', 'dining_room', 'wall', 'oven', 'newspaper', 'photoframe', 'check', 'toilet_paper', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'table']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": I walk in to the Kitchen. I find a plate, a sponge, and the dish soap. I find the faucet. I turn the faucet on. I pour water on the sponge and plate. I add dish soap to the sponge. I put the dish soap back. I scrub the plate. I rinse the plate. I place the plate on the drying rack. I rinse the sponge. I turn the faucet off.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <sponge>', '[GRAB] <sponge>', '[RINSE] <sponge>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[WASH] <plate>', '[RINSE] <plate>', '[FIND] <drying_rack>', '[PUTBACK] <plate> <plate>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'detergent', 'cat', 'drawing', 'floor', 'stamp', 'computer', 'ceilinglamp', 'mat', 'balanceball', 'picture', 'window', 'foundation', 'faucet', 'couch', 'pillow', 'wall', 'powersocket', 'oven', 'mouthwash', 'television', 'ice', 'centerpiece', 'towel_rack', 'mouse', 'kitchen_counter', 'dresser', 'table', 'dishwasher']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room. I see my cat on the couch. I go sit next to the cat on the couch. I pet the cat on the back.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'juice', 'toilet', 'freezer', 'knifeblock', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'mat', 'shower', 'dining_room', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'newspaper', 'food_food', 'curtain', 'doorjamb', 'cupboard', 'door', 'napkin', 'home_office', 'light', 'instrument_guitar']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'toilet', 'chair', 'bathroom', 'drawing', 'longboard', 'floor', 'ceiling', 'food_cake', 'mat', 'ceilinglamp', 'shower', 'deck_of_cards', 'balanceball', 'shoe-shine_kit', 'tvstand', 'candle', 'food_kiwi', 'wall', 'nightstand', 'powersocket', 'bench', 'food_cheese', 'novel', 'coffe_maker', 'curtain', 'towel_rack', 'home_office', 'dishwasher']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "Walk to the bathroom and wash with the shower.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>", "[WASH] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_pizza', 'tea', 'toilet', 'walllamp', 'nail_polish', 'cpuscreen', 'floor', 'phone', 'ceiling', 'keyboard', 'toaster', 'mat', 'oil', 'pillow', 'bowl', 'wall', 'powersocket', 'oven', 'photoframe', 'television', 'rag', 'doorjamb', 'headset', 'door', 'sink', 'home_office', 'light', 'bathroom_cabinet', 'coffee']\nTask:Turn light off\nresponse:\n{\n    \"thoughts\": Put the light switch into the down, or off, position.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <light>', '[FIND] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['food_salt', 'freezer', 'ceilingfan', 'clothes_underwear', 'bathroom_counter', 'phone', 'keyboard', 'computer', 'table_cloth', 'mop_bucket', 'tvstand', 'mousepad', 'candle', 'clothes_pants', 'basket_for_clothes', 'cup', 'bed', 'food_food', 'curtain', 'kettle', 'mouse', 'kitchen_counter', 'shoes', 'controller', 'clothes_dress', 'clothes_scarf', 'light', 'soap', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I need to get the dirty clothes basket.  Walk to bathroom.  Put all the clothes in washing machine.  Add soap.  Turn on machine.", "actions": ["[WALK] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'toilet', 'freezer', 'nail_polish', 'detergent', 'longboard', 'floor', 'ceiling', 'bedroom', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mousepad', 'couch', 'food_peanut_butter', 'form', 'wall', 'nightstand', 'oven', 'food_food', 'ice', 'after_shave', 'doorjamb', 'lighter', 'coffee_pot', 'mouse', 'spoon']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Put groceries into the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['food_salt', 'chair', 'walllamp', 'bills', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bathroom_counter', 'computer', 'tvstand', 'form', 'towel', 'wall', 'oven', 'photoframe', 'clothes_gloves', 'bed', 'food_food', 'doorjamb', 'towel_rack', 'dishrack', 'dresser', 'cupboard', 'door', 'light', 'table', 'colander', 'instrument_violin']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone and pick it up.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'chair', 'knifeblock', 'hanger', 'floor', 'toaster', 'stovefan', 'shoe-shine_kit', 'faucet', 'homework', 'food_noodles', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'cup', 'newspaper', 'coffee_filter', 'food_rice', 'console', 'mouse', 'kitchen_counter', 'dishrack', 'comb', 'cupboard', 'door', 'sink']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to the kitchen, walk to the cabinet, pic up a glass. carry the glass to the sink. turn on the water and fill up the glass. drink the water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'chair', 'freezer', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'balanceball', 'window', 'mousepad', 'wall', 'nightstand', 'bench', 'photoframe', 'mouthwash', 'doorjamb', 'mouse', 'pencil', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Find the computer.  Turn on the computer.  Open the e-mail program.  Create a new message.  Write the message.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'microwave', 'toilet', 'freezer', 'trashcan', 'drawing', 'knife', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'character', 'ceilinglamp', 'mat', 'shower', 'food_jam', 'dining_room', 'wall', 'bench', 'cutting_board', 'food_vegetable', 'photoframe', 'television', 'doorjamb', 'kitchen_counter', 'cupboard', 'door', 'bathtub', 'table', 'light']\nTask:Chop vegetables\nresponse:\n{\n    \"thoughts\": Cut up veggies with knife,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <knife>', '[GRAB] <knife>', '[FIND] <cutting_board>', '[GRAB] <cutting_board>', '[CLOSE] <kitchen_cabinet>', '[WALK] <table>', '[PUTBACK] <knife> <table>', '[PUTBACK] <cutting_board> <table>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <food_vegetable> <cutting_board>', '[GRAB] <knife>', '[CUT] <food_vegetable>', '[PUTOBJBACK] <knife>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'juice', 'food_salt', 'mop', 'jelly', 'ceilingfan', 'drawing', 'ceiling', 'character', 'desk', 'food_bacon', 'faucet', 'balanceball', 'powersocket', 'cutting_board', 'cup', 'coffe_maker', 'food_food', 'ice', 'tray', 'cards', 'towel_rack', 'hairbrush', 'cupboard', 'dresser', 'controller', 'door', 'napkin', 'home_office', 'vase']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I go looking for the cat in the living room where he usually sleeps. Once I find the cat I just pet him for a few minutes.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'ceilingfan', 'drawing', 'needle', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'clothes_socks', 'couch', 'pillow', 'wall', 'powersocket', 'food_cheese', 'bed', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk into the living room. Sit down on the sofa. Pick up the remote control. Turn on the television with the remote control. Put down the remote control. Watch the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['freezer', 'cat', 'drawing', 'knife', 'knifeblock', 'hanger', 'floor', 'bedroom', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'faucet', 'iron', 'window', 'mousepad', 'couch', 'form', 'dining_room', 'wall', 'oven', 'box', 'newspaper', 'bed', 'novel', 'toilet_paper', 'television', 'tray', 'sink', 'table']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room.  I locate the cat.  I walk to the couch and sit down beside the cat.  I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'drawing', 'needle', 'cpuscreen', 'bills', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'mousepad', 'tooth_paste', 'couch', 'pillow', 'towel', 'wall', 'nightstand', 'powersocket', 'television', 'coin', 'razor', 'doorjamb', 'centerpiece', 'mouse', 'door', 'light', 'board_game']\nTask:Re arrange office\nresponse:\n{\n    \"thoughts\": Walk in.Turn on light.Move furniture. Change the room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[PUSH] <desk>', '[WALK] <chair>', '[PUSH] <chair>', '[WALK] <couch>', '[PUSH] <couch>', '[WALK] <filing_cabinet>', '[PUSH] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['fork', 'beer', 'toilet', 'chair', 'ceilingfan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'folder', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'microphone', 'wall', 'bench', 'light_bulb', 'television', 'curtain', 'food_steak', 'sink', 'napkin', 'home_office', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into my living room, find the remote and sit on my sofa. I look at the remote and switch on the television. I look at the Television.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'detergent', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'food_cake', 'computer', 'mat', 'faucet', 'mousepad', 'bowl', 'wall', 'nightstand', 'oven', 'bench', 'box', 'check', 'coffee_filter', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Wipe down sink\nresponse:\n{\n    \"thoughts\": Use a cloth to wipe the edge of the sink. Then wipe the bowl of the sink as well.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[WIPE] <sink>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'mop', 'microwave', 'trashcan', 'food_egg', 'keyboard', 'mop_bucket', 'bookmark', 'mat', 'tvstand', 'mousepad', 'tooth_paste', 'food_noodles', 'food_jam', 'form', 'pillow', 'dining_room', 'food_kiwi', 'toy', 'keys', 'check', 'mouthwash', 'coffee_filter', 'shaving_cream', 'comb', 'maindoor', 'light', 'board_game', 'teeth', 'creditcard']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to living room, walk to the phone reach for the phone and pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'detergent', 'cat', 'drawing', 'bills', 'wallshelf', 'cpuscreen', 'phone', 'character', 'shower', 'food_orange', 'form', 'pillow', 'nightstand', 'photoframe', 'food_food', 'tray', 'lighter', 'shaving_cream', 'mouse', 'towel_rack', 'dishrack', 'laundry_detergent', 'spoon', 'bathroom_cabinet', 'light', 'home_office', 'soap', 'instrument_violin', 'shoe_rack']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Go into the living room and pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>', '[TOUCH] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['toilet', 'freezer', 'ceilingfan', 'dough', 'drawing', 'longboard', 'floor', 'ceiling', 'character', 'remote_control', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'couch', 'food_orange', 'bowl', 'wall', 'television', 'toothbrush_holder', 'tray', 'curtain', 'food_onion', 'mouse', 'door', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I push my body to be near to the table in front of me, without stand up. I move my right hand to near the TV remote control on the table. I take the TV remote control. I look the number on the remote control. I touch number 3 button. I put back the remote control to the table. I put my body back to the sofa", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[GRAB] <remote_control>", "[TOUCH] <remote_control>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'powersocket', 'oven', 'coffe_maker', 'television', 'coffee_filter', 'food_rice', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'dirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I like to set on my sofa.Turn on the tv. And watch Perry Mason. Monday's through Friday's.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['spectacles', 'drawing', 'knife', 'orchid', 'hanger', 'floor', 'keyboard', 'purse', 'bookshelf', 'mail', 'toaster', 'ceilinglamp', 'shower', 'balanceball', 'window', 'couch', 'clothes_pants', 'wall', 'oven', 'powersocket', 'television', 'coin', 'diary', 'cd', 'dresser', 'hairbrush', 'hair', 'conditioner', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'food_pizza', 'measuring_cup', 'microwave', 'nail_polish', 'bathroom', 'dough', 'bills', 'bathroom_counter', 'food_bacon', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'stereo', 'window', 'pillow', 'nightstand', 'bench', 'bed', 'clothes_gloves', 'keys', 'television', 'food_rice', 'lighter', 'cd', 'chessboard', 'table', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room. Turn on radio.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['toilet', 'cat', 'ceilingfan', 'food_butter', 'drawing', 'longboard', 'floor', 'ceiling', 'shampoo', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'couch', 'clothes_pants', 'wall', 'wall_clock', 'bench', 'television', 'curtain', 'kettle', 'towel_rack', 'hair', 'pajamas', 'maindoor', 'home_office', 'table', 'vase']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Changing channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[DROP] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'crayon', 'microwave', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'purse', 'toaster', 'faucet', 'shower', 'homework', 'pillow', 'wall', 'oven', 'water', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'home_office', 'soap', 'teeth']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to kitchen, turn on water, find soap, apply to hand, lather, rinse, dry hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[SCRUB] <soap>', '[PUTOBJBACK] <soap>', '[FIND] <water>', '[RINSE] <water>']\n}\n", "input": "\nScene_Objects:['scissors', 'microwave', 'dog', 'freezer', 'nail_polish', 'trashcan', 'knifeblock', 'floor', 'shower', 'picture', 'window', 'microphone', 'dining_room', 'broom', 'nightstand', 'photoframe', 'bed', 'clothes_gloves', 'lighter', 'console', 'pencil', 'kitchen_counter', 'cd', 'hair', 'maindoor', 'door', 'food_steak', 'home_office', 'table', 'sink']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to living room. Find book. Grab book. Open book. Read book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'chair', 'cat', 'drawing', 'floor', 'tea_bag', 'bookshelf', 'thread', 'toaster', 'ceilinglamp', 'stovefan', 'window', 'couch', 'food_orange', 'wall', 'food_food', 'television', 'oven_mitts', 'food_onion', 'curtain', 'after_shave', 'comb', 'napkin', 'home_office', 'table', 'dishwasher', 'bathroom_cabinet', 'laser_pointer', 'scrabble']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": go to the living room,find the sofa,sat down in the sofa,take some relaxation.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'measuring_cup', 'walllamp', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'food_noodles', 'couch', 'broom', 'wall', 'newspaper', 'light_bulb', 'novel', 'television', 'curtain', 'doorjamb', 'lighter', 'cupboard', 'table', 'light']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find remote control. Grab remote control. Walk to television. Look at television. Switch on remote control.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'blender', 'mop', 'bathroom', 'cat', 'trashcan', 'knifeblock', 'stamp', 'button', 'piano_bench', 'faucet', 'mousepad', 'bowl', 'nightstand', 'wall_clock', 'oven', 'wall', 'bench', 'photoframe', 'novel', 'food_food', 'coffe_maker', 'curtain', 'pencil', 'shoes', 'hairbrush', 'hair', 'instrument_guitar', 'dishwasher', 'tooth_paste']\nTask:Read book\nresponse:\n{\n    \"thoughts\": pick up  the book open the book and read,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['ceilingfan', 'trashcan', 'drawing', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'keyboard', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'brush', 'window', 'couch', 'wall', 'pasta', 'newspaper', 'television', 'wine', 'curtain', 'lighter', 'diary', 'hairbrush', 'dresser', 'controller', 'laundry_detergent', 'hairdryer', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I reach for the remote. I push the on button on the remote. I sit down on the couch and watch television.", "actions": ["[WALK] <home_office>", "[WALK] <button>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'freezer', 'walllamp', 'trashcan', 'food_carrot', 'cpuscreen', 'floor', 'ceiling', 'folder', 'bedroom', 'character', 'ceilinglamp', 'tvstand', 'faucet', 'dustpan', 'dining_room', 'wall', 'nightstand', 'oven', 'bench', 'photoframe', 'bed', 'food_food', 'television', 'food_donut', 'food_rice', 'doorjamb', 'shaving_cream', 'table', 'vase']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": walk to kitchen and seprate the groceries and arrange in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['sponge', 'beer', 'dog', 'freezer', 'trashcan', 'ceilingfan', 'food_carrot', 'bills', 'hanger', 'longboard', 'floor', 'ceiling', 'thread', 'mail', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'couch', 'wall', 'bench', 'powersocket', 'food_food', 'check', 'doorjamb', 'towel_rack', 'pencil', 'cupboard', 'door', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Find groceries. Grab groceries. Walk to refrigerator. Open refrigerator door. Place groceries into refrigerator. Close refrigerator door.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'cat', 'ceilingfan', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'dvd_player', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'foundation', 'oil', 'couch', 'towel', 'wall', 'pasta', 'powersocket', 'television', 'curtain', 'doorjamb', 'shoes', 'pajamas', 'maindoor', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'food_ice_cream', 'microwave', 'freezer', 'bathroom', 'trashcan', 'cpuscreen', 'bathroom_counter', 'ceiling', 'keyboard', 'tvstand', 'teeth', 'wall', 'oven', 'bench', 'toy', 'bed', 'photoframe', 'coffe_maker', 'food_food', 'television', 'food_donut', 'water', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'door', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the water, wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <water>", "[FIND] <water>", "[TURNTO] <water>", "[LOOKAT] <water>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'detergent', 'cat', 'drawing', 'bills', 'wallshelf', 'cpuscreen', 'phone', 'character', 'shower', 'food_orange', 'form', 'pillow', 'nightstand', 'photoframe', 'food_food', 'tray', 'lighter', 'shaving_cream', 'mouse', 'towel_rack', 'dishrack', 'laundry_detergent', 'spoon', 'bathroom_cabinet', 'light', 'home_office', 'soap', 'instrument_violin', 'shoe_rack']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Go into the living room and pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>', '[TOUCH] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'ceilingfan', 'milk', 'drawing', 'orchid', 'food_egg', 'ceiling', 'phone', 'keyboard', 'food_cake', 'remote_control', 'table_cloth', 'mop_bucket', 'window', 'food_kiwi', 'light_bulb', 'check', 'wooden_spoon', 'curtain', 'shaving_cream', 'pencil', 'hair', 'controller', 'clothes_dress', 'spoon', 'bag', 'home_office', 'dishwasher', 'vase']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'mop', 'chair', 'freezer', 'walllamp', 'knifeblock', 'clothes_underwear', 'floor', 'ceiling', 'character', 'stovefan', 'toaster', 'plate', 'ceilinglamp', 'faucet', 'pillow', 'wall', 'pasta', 'oven', 'cup', 'food_cheese', 'food_food', 'tray', 'curtain', 'kitchen_counter', 'sauce_pan', 'cupboard', 'sink', 'spoon', 'coffee']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <spoon>', '[GRAB] <spoon>', '[FIND] <sauce_pan>', '[PUTBACK] <spoon> <sauce_pan>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <plate>', '[PUTBACK] <pot> <plate>']\n}\n", "input": "\nScene_Objects:['crayon', 'microwave', 'toilet', 'freezer', 'ground_coffee', 'longboard', 'shampoo', 'character', 'thread', 'food_bacon', 'piano_bench', 'balanceball', 'stereo', 'window', 'picture', 'mousepad', 'pillow', 'toy', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'oven_mitts', 'colander', 'doorjamb', 'shaving_cream', 'dresser', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[WALK] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'chair', 'dog', 'toothbrush', 'drawing', 'bills', 'ceiling', 'folder', 'desk', 'bedroom', 'remote_control', 'button', 'clothes_jacket', 'stovefan', 'mat', 'faucet', 'dining_room', 'broom', 'nightstand', 'wall', 'cutting_board', 'novel', 'mouse', 'kitchen_counter', 'cd', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I pick up the remote. I press button on remote.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'trashcan', 'dough', 'orchid', 'knife', 'clothes_underwear', 'phone', 'tea_bag', 'envelope', 'mat', 'faucet', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bench', 'powersocket', 'basket_for_clothes', 'food_food', 'food_rice', 'tray', 'curtain', 'after_shave', 'doorjamb', 'pencil', 'laundry_detergent', 'bathroom_cabinet', 'table', 'light', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom and open the washing machine there. I take my laundry basket and put each item of laundry into the machine. Then I take the laundry detergent and pour some into the machine. Finally I close the machine and switch it on.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'chair', 'dog', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'bookshelf', 'computer', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'wall', 'powersocket', 'shelf', 'cup', 'doorjamb', 'console', 'mouse', 'shoes', 'dresser', 'bathtub', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I open the room door and press the light button on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'cd_player', 'bathroom', 'trashcan', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'longboard', 'character', 'bookmark', 'ceilinglamp', 'balanceball', 'window', 'couch', 'towel', 'wall', 'oven', 'bench', 'television', 'after_shave', 'towel_rack', 'door', 'sink', 'bathroom_cabinet', 'table', 'light', 'scrabble']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to the living room, switch on the CD player (then choose the music cd and insert into the player) and sit on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'floor', 'ceiling', 'food_chicken', 'stovefan', 'mat', 'toaster', 'shower', 'faucet', 'window', 'closetdrawer', 'couch', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'cupboard', 'door', 'sink']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": I walk in the kitchen, I open the fridge and remove the Chicken. I grab frying fan and put it on stove. I put chicken in the frying pan and I cook food.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_chicken>', '[GRAB] <food_chicken>', '[FIND] <pot>', '[GRAB] <pot>', '[FIND] <oven>', '[PUTBACK] <pot> <oven>', '[PUTBACK] <food_chicken> <oven>']\n}\n", "input": "\nScene_Objects:['sheets', 'fork', 'blender', 'food_ice_cream', 'food_butter', 'phone', 'character', 'bookshelf', 'brush', 'ceilinglamp', 'plate', 'shower', 'iron', 'stereo', 'toaster', 'oil', 'food_orange', 'pillow', 'cup', 'food_food', 'check', 'oven_mitts', 'food_donut', 'food_onion', 'after_shave', 'mouse', 'hairdryer', 'bag', 'bathroom_cabinet', 'home_office']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <home_office>", "[DROP] <plate>", "[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <home_office>", "[DROP] <cup>", "[WALK] <dining_room>", "[WALK] <fork>", "[FIND] <fork>", "[GRAB] <fork>", "[WALK] <home_office>", "[DROP] <fork>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'chair', 'bathroom', 'food_butter', 'phone', 'ceiling', 'bedroom', 'tea_bag', 'laptop', 'ceilinglamp', 'shower', 'iron', 'clothes_socks', 'closetdrawer', 'clothes_pants', 'pillow', 'wall', 'powersocket', 'standingmirror', 'photoframe', 'toothbrush_holder', 'food_donut', 'curtain', 'doorjamb', 'towel_rack', 'pencil', 'controller', 'door', 'table']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'toaster', 'laptop', 'plate', 'stovefan', 'shower', 'dish_soap', 'wall', 'bench', 'oven', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'kettle', 'pencil', 'kitchen_counter', 'cupboard', 'home_office', 'table', 'dishwasher', 'coffee']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Put the dishes on dinner table in the dishwasher and wash them", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[WALK] <table>", "[FIND] <pot>", "[GRAB] <pot>", "[WALK] <dishwasher>", "[PUTBACK] <pot> <dishwasher>", "[CLOSE] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'beer', 'bathroom', 'detergent', 'trashcan', 'cpuscreen', 'ceiling', 'bedroom', 'desk', 'keyboard', 'clothes_jacket', 'tvstand', 'couch', 'pillow', 'food_kiwi', 'oven', 'bench', 'coffe_maker', 'oven_mitts', 'towel_rack', 'pencil', 'cupboard', 'hair', 'clothes_dress', 'food_steak', 'napkin', 'vase', 'laser_pointer']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into the bedroom. I walk to stereo. I turn on stereo.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'toilet', 'chair', 'freezer', 'food_egg', 'cpuscreen', 'floor', 'desk', 'tea_bag', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'stovefan', 'laptop', 'picture', 'mousepad', 'dining_room', 'wall', 'nightstand', 'photoframe', 'bed', 'mouse', 'kitchen_counter', 'maindoor', 'home_office', 'light', 'clothes_skirt']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "walk to home office and switch on laptop and with the useof mouse and keyboard send mail", "actions": ["[WALK] <home_office>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'dry_pasta', 'filing_cabinet', 'scissors', 'kitchen_cabinet', 'envelope', 'faucet', 'dish_soap', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall_clock', 'bench', 'toy', 'cutting_board', 'razor', 'food_onion', 'doorjamb', 'towel_rack', 'kitchen_counter', 'pajamas', 'clothes_dress', 'bathtub', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'vase', 'light']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[GRAB] <dish_soap>', '[PUTIN] <dish_soap> <kitchen_cabinet>', '[CLOSE] <kitchen_cabinet>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'ground_coffee', 'dough', 'knifeblock', 'needle', 'floor', 'ceiling', 'phone', 'mop_bucket', 'toaster', 'stovefan', 'tvstand', 'mat', 'food_peanut_butter', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'check', 'coffee_filter', 'water', 'pajamas', 'light', 'dishwasher']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I walk into the kitchen. I find the coffee maker and open it. I find the coffee filter, grab it and put it into the coffee maker. I find ground coffee, grab it and put it into the coffee maker. I find some water, then I pour it into the coffee maker. I close the coffee maker, I plug it in and I switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'measuring_cup', 'freezer', 'walllamp', 'bathroom', 'hanger', 'phone', 'desk', 'bookshelf', 'ceilinglamp', 'shoe-shine_kit', 'balanceball', 'faucet', 'couch', 'nightstand', 'wall', 'bench', 'powersocket', 'clothes_gloves', 'check', 'television', 'food_onion', 'after_shave', 'curtain', 'shaving_cream', 'mouse', 'maindoor', 'sink', 'bag', 'home_office']\nTask:Open front door\nresponse:\n{\n    \"thoughts\": Walk over to the front door, turn the knob, and open the door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'dough', 'drawing', 'knifeblock', 'floor', 'desk', 'bookshelf', 'ceilinglamp', 'tvstand', 'stereo', 'balanceball', 'mousepad', 'couch', 'food_orange', 'wall', 'powersocket', 'box', 'cup', 'basket_for_clothes', 'coffe_maker', 'television', 'food_rice', 'tray', 'doorjamb', 'console', 'food_steak', 'sink', 'napkin', 'table']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "walk to entrance hall. find a table. pull table. walk to living rook. push table.", "actions": ["[WALK] <home_office>", "[WALK] <table>", "[FIND] <table>", "[PULL] <table>", "[WALK] <home_office>", "[WALK] <table>", "[PUSH] <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'freezer', 'drawing', 'bills', 'hanger', 'floor', 'ceiling', 'keyboard', 'computer', 'toaster', 'tvstand', 'floor_lamp', 'window', 'closetdrawer', 'couch', 'towel', 'wall', 'wall_clock', 'food_food', 'doorjamb', 'mouse', 'headset', 'shoes', 'dresser', 'controller', 'bathtub', 'sink', 'home_office', 'light']\nTask:Turn night light on\nresponse:\n{\n    \"thoughts\": I walk to the dresser and in the middle Io to small battery operated lamp and switch it on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <floor_lamp>', '[TURNTO] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['scissors', 'chair', 'dog', 'bathroom', 'knifeblock', 'knife', 'bathroom_counter', 'longboard', 'folder', 'bedroom', 'desk', 'keyboard', 'food_cake', 'computer', 'plate', 'tvstand', 'shoe-shine_kit', 'homework', 'dish_soap', 'couch', 'form', 'powersocket', 'food_food', 'oven_mitts', 'water', 'tray', 'doorjamb', 'mouse', 'dishrack', 'clothes_scarf']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen, fins soap, arrange dishes, wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <plate>", "[FIND] <plate>", "[FIND] <dish_soap>", "[FIND] <water>", "[GRAB] <plate>", "[GRAB] <dish_soap>", "[WASH] <plate>", "[RINSE] <plate>", "[DROP] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'character', 'video_game_console', 'bedroom', 'computer', 'button', 'table_cloth', 'tvstand', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'powersocket', 'box', 'video_game_controller', 'television', 'curtain', 'pencil', 'dresser', 'door', 'table']\nTask:Play games\nresponse:\n{\n    \"thoughts\": Turn on the television. Turn on the video game system. Insert desired game into gaming system. Follow directions on game to and push needed buttons on video  game remote controller.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <video_game_console>', '[SWITCHON] <video_game_console>', '[FIND] <video_game_controller>', '[GRAB] <video_game_controller>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'glue', 'tablelamp', 'microwave', 'walllamp', 'knife', 'knifeblock', 'longboard', 'floor', 'phone', 'computer', 'toaster', 'plate', 'shoe-shine_kit', 'dish_soap', 'mousepad', 'food_orange', 'wall', 'nightstand', 'oven', 'toy', 'newspaper', 'coffe_maker', 'tray', 'vacuum_cleaner', 'dresser', 'hairdryer', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the  kitchen. Open the dishwasher. Put plates into the dishwasher. Put dish soap into the dishwasher. Close the dishwasher. Turn on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'scissors', 'walllamp', 'needle', 'wallshelf', 'cpuscreen', 'hanger', 'bedroom', 'keyboard', 'bookmark', 'stovefan', 'shower', 'oil', 'dining_room', 'wall', 'powersocket', 'bench', 'love_seat', 'standingmirror', 'novel', 'coffe_maker', 'check', 'after_shave', 'towel_rack', 'mouse', 'pencil', 'kitchen_counter', 'maindoor', 'bathtub', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'floor_lamp', 'electric_shaver', 'wall', 'nightstand', 'powersocket', 'bed', 'doorjamb', 'shoes', 'hairbrush', 'door', 'light', 'table', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to wall, find light switch, turn on lights", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[WALK] <bedroom>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'newspaper', 'television', 'curtain', 'doorjamb', 'door', 'home_office', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room. look at remote control. grab remote control. plug in television. switch on television. lie in sofa. point at remote control. watch television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[PLUGIN] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[LIE] <couch>', '[TURNTO] <remote_control>', '[POINTAT] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'measuring_cup', 'microwave', 'freezer', 'ceilingfan', 'milk', 'floor', 'ceiling', 'character', 'bedroom', 'computer', 'button', 'toaster', 'faucet', 'picture', 'wall', 'nightstand', 'oven', 'photoframe', 'clothes_gloves', 'coffe_maker', 'food_food', 'curtain', 'centerpiece', 'food_sugar', 'vacuum_cleaner', 'chessboard', 'coffee_cup', 'dishwasher', 'light']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[LOOKAT] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <milk>", "[GRAB] <milk>", "[PUTBACK] <milk> <coffe_maker>", "[FIND] <food_sugar>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <coffe_maker>", "[SWITCHON] <coffe_maker>", "[FIND] <button>", "[PUSH] <button>", "[GRAB] <coffee_cup>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'band-aids', 'freezer', 'nail_polish', 'bathroom', 'cat', 'orchid', 'cpuscreen', 'phone', 'character', 'keyboard', 'mat', 'faucet', 'wall', 'oven', 'bed', 'keys', 'coffe_maker', 'mouthwash', 'food_onion', 'curtain', 'centerpiece', 'console', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'home_office']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'dog', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'folder', 'desk', 'bookshelf', 'keyboard', 'thread', 'computer', 'mail', 'bookmark', 'mousepad', 'wall', 'nightstand', 'photoframe', 'novel', 'toilet_paper', 'tray', 'towel_rack', 'mouse', 'door', 'light']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find a desk and grab a book. Walk to the chair, sit in the chair and read a book.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <desk>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'dog', 'bathroom', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'computer', 'ceilinglamp', 'window', 'mousepad', 'couch', 'towel', 'wall', 'cup', 'television', 'curtain', 'doorjamb', 'dresser', 'cupboard', 'door', 'sink', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room and turn on the light. I walk to the couch then sit on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'microwave', 'toilet', 'jelly', 'drawing', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'ceilinglamp', 'deck_of_cards', 'stereo', 'shower', 'window', 'couch', 'candle', 'microphone', 'wall', 'powersocket', 'box', 'photoframe', 'television', 'tray', 'towel_rack', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'home_office']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'chair', 'bathroom', 'knifeblock', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'character', 'bedroom', 'desk', 'stovefan', 'ceilinglamp', 'mat', 'faucet', 'homework', 'mousepad', 'form', 'powersocket', 'bench', 'light_bulb', 'check', 'television', 'tray', 'curtain', 'chessboard', 'cupboard', 'home_office']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": I walk to the home office, find where the phone is and lift the reciever.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'mousepad', 'couch', 'dining_room', 'wall', 'wall_clock', 'powersocket', 'oven', 'television', 'toothbrush_holder', 'curtain', 'doorjamb', 'console', 'dresser', 'cupboard', 'home_office', 'table', 'clothes_skirt', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch the television", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[SWITCHON] <television>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'deck_of_cards', 'faucet', 'mousepad', 'pillow', 'wall', 'oven', 'box', 'cup', 'toilet_paper', 'rag', 'kitchen_counter', 'cupboard', 'sink', 'bag', 'bathroom_cabinet', 'bathtub', 'table', 'home_office']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk the kitchen and walk up to the kitchen cabinet. I open the cabinet and take out a cup. I walk to sink and I turn the faucet on using the cold water handle. I put my cup under the running water until it is full. I turn off the cold water handle and bring the cup to my mouth and take a drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <cup>', '[GRAB] <cup>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[WALK] <sink>', '[FIND] <faucet>', '[PUSH] <faucet>', '[PUTBACK] <cup> <sink>', '[GRAB] <cup>', '[SWITCHOFF] <faucet>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'ceilingfan', 'drawing', 'longboard', 'floor', 'ceiling', 'phone', 'folder', 'character', 'computer', 'food_bacon', 'toaster', 'brush', 'ceilinglamp', 'balanceball', 'tvstand', 'stereo', 'oil', 'couch', 'wall', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'rag', 'clothes_dress', 'bathroom_cabinet', 'dishwasher', 'coffee', 'shoe_rack']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "I find the coffee machine, fill it with water and coffee then press the button to make the coffee", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[WALK] <coffee>", "[FIND] <coffee>", "[GRAB] <coffee>", "[POUR] <coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'drawing', 'knife', 'orchid', 'hanger', 'floor', 'keyboard', 'purse', 'bookshelf', 'mail', 'toaster', 'ceilinglamp', 'shower', 'balanceball', 'window', 'couch', 'clothes_pants', 'wall', 'oven', 'powersocket', 'television', 'coin', 'diary', 'cd', 'dresser', 'hairbrush', 'hair', 'conditioner', 'table', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'crayon', 'detergent', 'ceilingfan', 'food_carrot', 'milk', 'drawing', 'knifeblock', 'food_egg', 'floor', 'phone', 'desk', 'bookshelf', 'envelope', 'ceilinglamp', 'window', 'mousepad', 'couch', 'wall', 'food_cheese', 'novel', 'television', 'food_rice', 'after_shave', 'towel_rack', 'kitchen_counter', 'comb', 'table', 'dishwasher', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room and find the sofa. I sit on the sofa and get comfortable.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tea', 'microwave', 'chair', 'freezer', 'walllamp', 'milk', 'knifeblock', 'knife', 'drawing', 'floor', 'ceiling', 'shampoo', 'desk', 'bookshelf', 'stovefan', 'toaster', 'faucet', 'food_jam', 'pillow', 'wall', 'oven', 'powersocket', 'cup', 'photoframe', 'tray', 'kitchen_counter', 'cupboard', 'controller', 'sink']\nTask:Get something to drink\nresponse:\n{\n    \"thoughts\": Walk into the kitchen and to the cupboard to get a cup, walk to the fridge to look for a beverage to enjoy, open the freezer and get ice, pour beverage and enjoy,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[CLOSE] <cupboard>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <milk>', '[GRAB] <milk>', '[CLOSE] <freezer>', '[POUR] <milk> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['pot', 'ceilingfan', 'drawing', 'food_egg', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_jam', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'television', 'food_rice', 'curtain', 'kitchen_counter', 'hair', 'home_office', 'table', 'light', 'dishwasher', 'dirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go to the living room where the TV is located. I plug in the TV to the outlet. I search for the remote control. I point the remote control directly to the TV, turn it on and push the number of the channel that I want.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'food_cake', 'envelope', 'computer', 'ceilinglamp', 'window', 'mousepad', 'clothes_socks', 'closetdrawer', 'wall', 'toy', 'cup', 'coffe_maker', 'doorjamb', 'mouse', 'dresser', 'door', 'light', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I turn on computer. I use keyboard to type. I open browser. I search.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'knifeblock', 'needle', 'bills', 'drawing', 'phone', 'mat', 'deck_of_cards', 'shoe-shine_kit', 'ceilinglamp', 'electric_shaver', 'window', 'clothes_socks', 'couch', 'dining_room', 'nightstand', 'wall', 'bench', 'cup', 'bed', 'mouthwash', 'coffee_filter', 'rag', 'towel_rack', 'hairbrush', 'dresser', 'clothes_dress', 'bathroom_cabinet', 'laser_pointer']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk to phone, grab phone, pick up phone, put to ear", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'walllamp', 'bathroom', 'knifeblock', 'food_egg', 'food_bacon', 'brush', 'mat', 'deck_of_cards', 'balanceball', 'tvstand', 'ceilinglamp', 'picture', 'couch', 'wall', 'tape', 'cup', 'toilet_paper', 'food_food', 'video_game_controller', 'water', 'cupboard', 'bathroom_cabinet', 'table', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'chair', 'ceilingfan', 'dough', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'folder', 'bedroom', 'character', 'ceilinglamp', 'mat', 'balanceball', 'mousepad', 'wall', 'nightstand', 'box', 'check', 'television', 'doorjamb', 'pencil', 'vacuum_cleaner', 'dresser', 'cupboard', 'conditioner', 'door', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the kitchen, I find the light switch, and then i switch on the lights in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'toilet', 'jelly', 'freezer', 'ground_coffee', 'food_carrot', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'button', 'mat', 'shower', 'picture', 'dustpan', 'food_orange', 'form', 'nightstand', 'wall', 'powersocket', 'food_cheese', 'photoframe', 'light_bulb', 'food_food', 'doorjamb', 'cd', 'door', 'light']\nTask:Added meat to freezer\nresponse:\n{\n    \"thoughts\": After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'beer', 'tea', 'freezer', 'walllamp', 'nail_polish', 'bathroom', 'needle', 'cpuscreen', 'keyboard', 'mail', 'computer', 'stovefan', 'ceilinglamp', 'iron', 'clothes_pants', 'food_orange', 'pillow', 'dining_room', 'bench', 'bed', 'coffe_maker', 'television', 'doorjamb', 'pencil', 'maindoor', 'napkin', 'instrument_violin']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the home office, find where the phone is and lift the reciever.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dog', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'door', 'bathtub', 'sink', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room and go to sofa and turn on tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[WALK] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[WALK] <couch>', '[LIE] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[DROP] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'band-aids', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'food_bacon', 'mat', 'tvstand', 'faucet', 'window', 'clothes_pants', 'wall', 'powersocket', 'after_shave', 'doorjamb', 'cd', 'dresser', 'door', 'bathroom_cabinet', 'light', 'dishwasher', 'vase']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Turning on lights", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <bathroom>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_bread', 'freezer', 'food_carrot', 'milk', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'piano_bench', 'ceilinglamp', 'mousepad', 'garbage_can', 'clothes_hat', 'food_orange', 'food_dessert', 'wall', 'powersocket', 'food_vegetable', 'food_cheese', 'doorjamb', 'lighter', 'mouse', 'coffee', 'door', 'light', 'colander', 'teeth']\nTask:Rotate stock in refrigerator\nresponse:\n{\n    \"thoughts\": Throw out expired items, put older items in front.,\n    \"actions\": ['[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_bread>', '[GRAB] <food_bread>', '[TURNTO] <food_bread>', '[LOOKAT] <food_bread>', '[PUTOBJBACK] <food_bread>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[FIND] <garbage_can>', '[PUTBACK] <food_cheese> <garbage_can>', '[FIND] <food_cheese>', '[GRAB] <food_cheese>', '[TURNTO] <food_cheese>', '[LOOKAT] <food_cheese>', '[PUTOBJBACK] <food_cheese>', '[FIND] <food_dessert>', '[TURNTO] <food_dessert>', '[LOOKAT] <food_dessert>', '[PUSH] <food_dessert>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[TURNTO] <food_vegetable>', '[LOOKAT] <food_vegetable>', '[PUTBACK] <food_vegetable> <garbage_can>', '[FIND] <milk>', '[GRAB] <milk>', '[TURNTO] <milk>', '[LOOKAT] <milk>', '[PUTOBJBACK] <milk>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'knife', 'cpuscreen', 'floor', 'ceiling', 'desk', 'remote_control', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'dining_room', 'wall', 'tape', 'photoframe', 'television', 'tray', 'curtain', 'kitchen_counter', 'vacuum_cleaner', 'sink', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'microwave', 'walllamp', 'drawing', 'knifeblock', 'cpuscreen', 'orchid', 'wallshelf', 'needle', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'faucet', 'wall', 'oven', 'powersocket', 'bench', 'after_shave', 'curtain', 'centerpiece', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'soap']\nTask:Wipe down counter\nresponse:\n{\n    \"thoughts\": Clean counter,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <rag>', '[SQUEEZE] <rag>', '[SWITCHOFF] <faucet>', '[WALK] <kitchen_counter>', '[WIPE] <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'bookshelf', 'keyboard', 'computer', 'food_bacon', 'stovefan', 'mousepad', 'wall', 'nightstand', 'powersocket', 'photoframe', 'novel', 'check', 'television', 'food_rice', 'mouse', 'dresser', 'light', 'dishwasher', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Walk to bookcase in office. Take book from desk. Take book over to chair. Sit down in the chair. Open book and then read book.", "actions": ["[WALK] <home_office>", "[WALK] <desk>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'freezer', 'walllamp', 'bathroom', 'drawing', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'brush', 'tvstand', 'stereo', 'window', 'closetdrawer', 'towel', 'wall', 'powersocket', 'tape', 'food_food', 'television', 'curtain', 'doorjamb', 'vacuum_cleaner', 'dresser', 'door', 'clothes_scarf', 'dishwasher', 'board_game']\nTask:vacuum carpet\nresponse:\n{\n    \"thoughts\": I bring my vacuum cleaner into the living room.  I plug it into the wall.  I press the on switch on the vacuum cleaner.  I push the vacuum cleaner back and forth over the length and width of the carpet until I have vacuumed the entire carpet.  I turn the vacuum cleaner off.  I take the vacuum cleaner back to the hall closet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dresser>', '[OPEN] <dresser>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PLUGIN] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[PULL] <vacuum_cleaner>', '[PUSH] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>', '[PLUGOUT] <vacuum_cleaner>', '[WALK] <dresser>', '[PUTIN] <vacuum_cleaner> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['juice', 'toilet', 'chair', 'blow_dryer', 'food_carrot', 'orchid', 'knifeblock', 'cpuscreen', 'desk', 'clothes_jacket', 'toaster', 'deck_of_cards', 'tvstand', 'dining_room', 'bench', 'basket_for_clothes', 'novel', 'television', 'after_shave', 'coffee_pot', 'mouse', 'cupboard', 'clothes_dress', 'hairdryer', 'home_office', 'colander', 'clothes_shirt', 'washing_machine', 'coffee', 'shoe_rack']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Find the washing machine. Open the washing machine. Find the basket for clothes. Find the clothes_dress. Grab the clothes_dress. Put the clothes_dress into the washing machine. Find the clothes_shirt. Grab the clothes_shirt. Put the clothes_shirt into the washing machine. Close the washing machine. Switch on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <basket_for_clothes>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'juice', 'microwave', 'freezer', 'bills', 'floor', 'ceiling', 'desk', 'tea_bag', 'computer', 'ceilinglamp', 'plate', 'window', 'mousepad', 'couch', 'dining_room', 'broom', 'wall', 'powersocket', 'novel', 'food_food', 'mouthwash', 'television', 'curtain', 'rag', 'towel_rack', 'kitchen_counter', 'comb', 'door', 'laser_pointer']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['food_bread', 'chair', 'detergent', 'food_carrot', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'bookmark', 'toaster', 'balanceball', 'wall', 'powersocket', 'photoframe', 'coffee_filter', 'tray', 'doorjamb', 'cards', 'mouse', 'shaving_cream', 'shoes', 'dresser', 'cupboard', 'door', 'napkin', 'home_office', 'light', 'vase']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on light in an area", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'tablelamp', 'microwave', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'food_bacon', 'stovefan', 'mat', 'shower', 'couch', 'pillow', 'wall', 'oven', 'powersocket', 'food_food', 'television', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'door', 'light']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": go to the kitchen and cook some thing to eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <oven>', '[OPEN] <oven>', '[FIND] <food_pizza>', '[GRAB] <food_pizza>', '[PUTIN] <food_pizza> <oven>', '[SWITCHON] <oven>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_bread', 'microwave', 'dog', 'chair', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'oven', 'newspaper', 'check', 'television', 'coffee_filter', 'oven_mitts', 'coin', 'curtain', 'lighter', 'maindoor', 'napkin', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control", "actions": ["[WALK] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[TOUCH] <button>", "[TURNTO] <television>", "[WATCH] <television>", "[PUTOBJBACK] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'mop', 'chair', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'tea_bag', 'bookshelf', 'computer', 'toaster', 'faucet', 'mousepad', 'closetdrawer', 'clothes_hat', 'wall', 'powersocket', 'bed', 'standingmirror', 'television', 'doorjamb', 'cards', 'mouse', 'dresser', 'hair', 'light', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.,\n    \"actions\": ['[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_salt', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'desk', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'clothes_pants', 'dining_room', 'wall', 'toy', 'basket_for_clothes', 'bed', 'television', 'coin', 'coffee_filter', 'food_rice', 'curtain', 'cupboard', 'controller', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Watch TV in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'microwave', 'walllamp', 'food_butter', 'knifeblock', 'drawing', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'tvstand', 'balanceball', 'picture', 'shower', 'oil', 'food_peanut_butter', 'pillow', 'wall', 'oven', 'television', 'coin', 'diary', 'mouse', 'cupboard', 'bag', 'bathroom_cabinet', 'vase', 'shoe_rack']\nTask:Make popcorn\nresponse:\n{\n    \"thoughts\": Lay popcorn bag flat inside microwave and press popcorn button to start.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <microwave>', '[TURNTO] <microwave>', '[OPEN] <microwave>', '[PUTIN] <food_food> <microwave>', '[CLOSE] <microwave>', '[SWITCHON] <microwave>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'keyboard', 'computer', 'food_bacon', 'alcohol', 'ceilinglamp', 'pillow', 'wall', 'powersocket', 'bed', 'curtain', 'doorjamb', 'kettle', 'mouse', 'dresser', 'cupboard', 'door', 'maindoor', 'light', 'creditcard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I open the room door and press the light button on", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'trashcan', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'ceilinglamp', 'mat', 'faucet', 'shower', 'couch', 'towel', 'wall', 'bench', 'oven', 'photoframe', 'bed', 'mouse', 'kitchen_counter', 'dresser', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[FIND] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['fork', 'food_bread', 'blender', 'food_salt', 'toilet', 'jelly', 'freezer', 'trashcan', 'clothes_underwear', 'longboard', 'phone', 'character', 'keyboard', 'thread', 'clothes_jacket', 'mop_bucket', 'stereo', 'electric_shaver', 'window', 'pillow', 'dining_room', 'powersocket', 'cup', 'television', 'tray', 'curtain', 'shaving_cream', 'kitchen_counter', 'laundry_detergent', 'home_office']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "turn radio on, listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'tvstand', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'food_jam', 'wall', 'nightstand', 'oven', 'newspaper', 'food_food', 'vacuum_cleaner', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'laptop', 'faucet', 'mousepad', 'candle', 'microphone', 'food_orange', 'wall', 'photoframe', 'check', 'mouthwash', 'after_shave', 'doorjamb', 'cupboard', 'pajamas', 'conditioner', 'door', 'sink', 'bathroom_cabinet', 'light', 'soap', 'clothes_shirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go into the bathroom and turn on the light. I walk to the sink and turn on the water. I grab the soap and wash my hands", "actions": ["[WALK] <bathroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[PUTOBJBACK] <soap>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'trashcan', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'mousepad', 'wall', 'bench', 'photoframe', 'novel', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'shaving_cream', 'door', 'bathroom_cabinet', 'light', 'home_office']\nTask:Read\nresponse:\n{\n    \"thoughts\": We pick out a book from the bookshelf and read it.,\n    \"actions\": ['[WALK] <novel>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'blender', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'orchid', 'food_egg', 'kitchen_cabinet', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'ceilinglamp', 'toaster', 'faucet', 'wall', 'oven', 'food_cheese', 'food_food', 'food_donut', 'food_rice', 'wine', 'tray', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into kitchen. Open fridge. Take food from fridge. Close fridge. Put food on kitchen counter. Prepare food. Turn on stove. Open kitchen cabinet. Take pot. Close cabinet. Put food in pot. Put pot on stove. Cook food.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <kitchen_counter>", "[PUTBACK] <food_food> <kitchen_counter>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[FIND] <oven>", "[PUTBACK] <pot> <oven>", "[GRAB] <food_food>", "[PUTBACK] <food_food> <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'microwave', 'freezer', 'walllamp', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'alcohol', 'stovefan', 'toaster', 'tvstand', 'couch', 'wall', 'oven', 'cutting_board', 'toy', 'food_cheese', 'novel', 'coffe_maker', 'food_food', 'television', 'tray', 'shaving_cream', 'cards', 'kitchen_counter', 'shoes', 'light']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": put away groceries in fridge. open the door, put on shelf, close the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'blow_dryer', 'bathroom', 'detergent', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'shampoo', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'shower', 'food_jam', 'towel', 'dining_room', 'wall', 'powersocket', 'after_shave', 'doorjamb', 'centerpiece', 'door', 'maindoor', 'light', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to room. I enter room. I find where the lightswitch is. I turn on the lights with the lightswitch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[TURNTO] <light>", "[LOOKAT] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'blender', 'food_ice_cream', 'microwave', 'chair', 'walllamp', 'bathroom', 'trashcan', 'milk', 'floor', 'bedroom', 'computer', 'toaster', 'plate', 'shoe-shine_kit', 'oil', 'wall', 'newspaper', 'coffe_maker', 'tray', 'centerpiece', 'console', 'pencil', 'kitchen_counter', 'cupboard', 'spoon', 'dishwasher', 'laser_pointer', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I take plate. I open the dishwasher. I put plate in the dishwasher. I switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_bread', 'walllamp', 'bathroom', 'cpuscreen', 'kitchen_cabinet', 'floor', 'tea_bag', 'toaster', 'mat', 'stereo', 'balanceball', 'dish_soap', 'towel', 'wall', 'oven', 'bed', 'light_bulb', 'novel', 'coffe_maker', 'check', 'coin', 'food_donut', 'tray', 'doorjamb', 'towel_rack', 'sink', 'dishwasher', 'vase', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I open the dishwasher. I open the kitchen cabinet. I reach inside the cabinet and grab the box of dish soap. I add some soap to the open dishwasher. I put the box of dish soap back in the kitchen cabinet. I close the kitchen cabinet and then I close the dishwasher. I turn the dishwasher on and wait.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTIN] <dish_soap> <kitchen_cabinet>", "[CLOSE] <kitchen_cabinet>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'tablelamp', 'blender', 'walllamp', 'dough', 'knifeblock', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'bookshelf', 'clothes_jacket', 'faucet', 'balanceball', 'pillow', 'wall', 'oven', 'food_food', 'wooden_spoon', 'curtain', 'kitchen_counter', 'headset', 'dishrack', 'cupboard', 'maindoor', 'sink', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Washing Hands,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[TOUCH] <soap>', '[PUTOBJBACK] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sheets', 'drawing', 'cpuscreen', 'ceiling', 'desk', 'bookmark', 'mat', 'plate', 'stereo', 'deck_of_cards', 'mousepad', 'couch', 'clothes_hat', 'clothes_pants', 'bench', 'clothes_gloves', 'video_game_controller', 'curtain', 'towel_rack', 'cd', 'dishrack', 'chessboard', 'pajamas', 'laundry_detergent', 'maindoor', 'door', 'sink', 'bathroom_cabinet', 'clothes_shirt']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Find the radio; switch on the radio and listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'juice', 'walllamp', 'food_carrot', 'knife', 'orchid', 'cpuscreen', 'phone', 'dvd_player', 'desk', 'mop_bucket', 'toaster', 'laptop', 'tvstand', 'closetdrawer', 'food_jam', 'pillow', 'dining_room', 'food_food', 'television', 'oven_mitts', 'water', 'after_shave', 'pencil', 'hairbrush', 'clothes_dress', 'door', 'table', 'dishwasher', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'walllamp', 'drawing', 'bills', 'knifeblock', 'hanger', 'floor', 'phone', 'ceilinglamp', 'tvstand', 'faucet', 'iron', 'window', 'mousepad', 'couch', 'food_peanut_butter', 'clothes_hat', 'clothes_pants', 'wall', 'basket_for_clothes', 'clothes_gloves', 'check', 'television', 'after_shave', 'headset', 'maindoor', 'spoon', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "allow sofa to support weight in order to decrease gravity induced entropy", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'chair', 'walllamp', 'dough', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shoe-shine_kit', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'novel', 'television', 'doorjamb', 'diary', 'shoes', 'dresser', 'door', 'clothes_shirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room. I sit on the couch. I pick up the remote control. I push the power button.  I push the guide button and look for my favorite show \"The Middle.\"  I click that channel and enjoy my show.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUSH] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'food_bread', 'tablelamp', 'microwave', 'freezer', 'ceilingfan', 'food_carrot', 'orchid', 'longboard', 'bookshelf', 'mail', 'food_bacon', 'brush', 'ceilinglamp', 'tvstand', 'stereo', 'balanceball', 'clothes_pants', 'pillow', 'wall', 'bench', 'powersocket', 'curtain', 'mouse', 'controller', 'maindoor', 'home_office', 'board_game', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the radio. Turn on the radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'keys', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'light', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I want to relax on the sofa to some relaxing music in the living room.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <stereo>', '[PLUGIN] <stereo>', '[SWITCHON] <stereo>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['sheets', 'freezer', 'ground_coffee', 'bathroom', 'ceilingfan', 'drawing', 'clothes_underwear', 'floor', 'food_cake', 'toaster', 'mousepad', 'dining_room', 'oven', 'tape', 'box', 'photoframe', 'coin', 'curtain', 'shaving_cream', 'coffee_pot', 'headset', 'dishrack', 'cupboard', 'sink', 'clothes_scarf', 'home_office', 'dishwasher', 'soap', 'board_game', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.", "actions": ["[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'detergent', 'ceilingfan', 'hanger', 'bathroom_counter', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'food_cake', 'clothes_jacket', 'button', 'laptop', 'tvstand', 'mat', 'dustpan', 'food_food', 'after_shave', 'towel_rack', 'pencil', 'kitchen_counter', 'laundry_detergent', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'table', 'vase', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'band-aids', 'freezer', 'nail_polish', 'bathroom', 'cat', 'orchid', 'cpuscreen', 'phone', 'character', 'keyboard', 'mat', 'faucet', 'wall', 'oven', 'bed', 'keys', 'coffe_maker', 'mouthwash', 'food_onion', 'curtain', 'centerpiece', 'console', 'towel_rack', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'home_office']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'tea', 'bathroom_counter', 'floor', 'phone', 'keyboard', 'table_cloth', 'mat', 'stereo', 'window', 'foundation', 'mousepad', 'couch', 'clothes_hat', 'pillow', 'wall', 'nightstand', 'bench', 'keys', 'toilet_paper', 'food_food', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'vacuum_cleaner', 'comb', 'maindoor', 'bathtub', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": move to living room, locate sofa, move to sofa, sit on sofa, lie on sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['food_salt', 'food_oatmeal', 'microwave', 'freezer', 'ceilingfan', 'food_carrot', 'orchid', 'drawing', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'longboard', 'bathroom_counter', 'stovefan', 'ceilinglamp', 'shower', 'food_cereal', 'wall', 'powersocket', 'nightstand', 'food_vegetable', 'food_food', 'doorjamb', 'shaving_cream', 'hairbrush', 'spoon', 'dishwasher', 'vase', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen, find groceries in cabinet and move them to fridge and close fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <food_vegetable>", "[GRAB] <food_vegetable>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_vegetable> <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[PUTIN] <food_carrot> <freezer>", "[FIND] <food_cereal>", "[GRAB] <food_cereal>", "[PUTIN] <food_cereal> <freezer>", "[FIND] <food_oatmeal>", "[GRAB] <food_oatmeal>", "[PUTIN] <food_oatmeal> <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[FIND] <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'button', 'mat', 'balanceball', 'window', 'candle', 'wall', 'powersocket', 'photoframe', 'novel', 'coffee_filter', 'ice', 'doorjamb', 'kettle', 'console', 'door', 'bathroom_cabinet', 'light', 'home_office', 'table', 'soap']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.,\n    \"actions\": ['[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'food_salt', 'walllamp', 'bathroom', 'trashcan', 'ceilingfan', 'floor', 'keyboard', 'computer', 'brush', 'window', 'candle', 'food_jam', 'pillow', 'food_kiwi', 'wall', 'pasta', 'wall_clock', 'bench', 'bed', 'wooden_spoon', 'curtain', 'towel_rack', 'mouse', 'pencil', 'maindoor', 'light', 'table', 'coffee']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Enter bedroom, turn on the light, and shut door. Change into night clothes. Turn off light. Turn down comforter, climb into bed, lay head on pillow, and cover up with comforter. Close eyes and drift off to sleep.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>", "[FIND] <sheets>", "[PULL] <sheets>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'shampoo', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'closetdrawer', 'couch', 'pillow', 'wall', 'child', 'photoframe', 'food_food', 'television', 'after_shave', 'doorjamb', 'dresser', 'sink', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Turn on tv.  Sit with kids.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[FIND] <child>', '[GREET] <child>', '[FIND] <child>', '[GREET] <child>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'remote_control', 'alcohol', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'couch', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'curtain', 'rag', 'doorjamb', 'dresser', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to living room, sit on couch, get remote, switch on tv, watch tv", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'detergent', 'knifeblock', 'cpuscreen', 'longboard', 'folder', 'character', 'bookshelf', 'table_cloth', 'toaster', 'stereo', 'picture', 'food_orange', 'nightstand', 'wall_clock', 'tape', 'light_bulb', 'coffe_maker', 'ice', 'tray', 'curtain', 'doorjamb', 'cd', 'shoes', 'dresser', 'door', 'sink', 'light', 'dishwasher', 'teeth']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find a radio there and switch it on to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_bread', 'freezer', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'bookshelf', 'thread', 'alcohol', 'ceilinglamp', 'shoe-shine_kit', 'balanceball', 'wall', 'oven', 'food_food', 'coffe_maker', 'toilet_paper', 'check', 'food_rice', 'tray', 'towel_rack', 'pencil', 'kitchen_counter', 'napkin', 'dishwasher', 'clothes_shirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Open the fridge, and put groceries in the fridge. Close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'dry_pasta', 'tablelamp', 'walllamp', 'detergent', 'milk', 'orchid', 'floor', 'ceiling', 'button', 'toaster', 'faucet', 'wall', 'box', 'basket_for_clothes', 'keys', 'coffe_maker', 'food_food', 'curtain', 'food_sugar', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'bathtub', 'sink', 'home_office', 'coffee_cup', 'dishwasher', 'clothes_skirt']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": walk to kitchen, look at the coffee maker, open the coffee maker, find milk, suger and water, mix that all, put into coffee maker, switch on the coffee maker, wait for few mins, push the coffee button in coffee maker, we get coffee in coffee cup, drink coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[TURNTO] <coffe_maker>', '[LOOKAT] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <milk>', '[GRAB] <milk>', '[PUTBACK] <milk> <coffe_maker>', '[FIND] <food_sugar>', '[FIND] <coffee_cup>', '[GRAB] <coffee_cup>', '[PUTBACK] <coffee_cup> <coffe_maker>', '[SWITCHON] <coffe_maker>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <coffee_cup>', '[DRINK] <coffee_cup>']\n}\n", "input": "\nScene_Objects:['fork', 'water_glass', 'food_pizza', 'tablelamp', 'drawing', 'knifeblock', 'needle', 'bathroom_counter', 'floor', 'desk', 'purse', 'food_bacon', 'ceilinglamp', 'mat', 'faucet', 'picture', 'window', 'couch', 'wall', 'cup', 'check', 'toilet_paper', 'television', 'water', 'tray', 'hairbrush', 'cupboard', 'controller', 'clothes_scarf', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "GOING TO DRINK WATER IN KITCHEN", "actions": ["[WALK] <dining_room>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'dvd_player', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'mousepad', 'wall', 'nightstand', 'powersocket', 'coffe_maker', 'check', 'television', 'wine', 'doorjamb', 'mouse', 'cupboard', 'controller', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to my desk. I turn on the computer. I open a web page and look at it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'crayon', 'freezer', 'bathroom', 'trashcan', 'dough', 'knifeblock', 'orchid', 'floor', 'ceiling', 'clothes_jacket', 'brush', 'ceilinglamp', 'mat', 'balanceball', 'clothes_socks', 'couch', 'microphone', 'form', 'wall', 'food_food', 'check', 'oven_mitts', 'wine', 'mouse', 'cupboard', 'food_steak', 'spoon', 'table', 'tooth_paste']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Go kitchen and make something for eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'mat', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'powersocket', 'check', 'television', 'wine', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'light', 'table', 'clothes_skirt']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": enter office, find computer, turn it on,,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TOUCH] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tea', 'microwave', 'band-aids', 'drawing', 'cpuscreen', 'hanger', 'floor', 'computer', 'ceilinglamp', 'balanceball', 'window', 'oil', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'coffe_maker', 'toilet_paper', 'television', 'kitchen_counter', 'chessboard', 'home_office', 'table', 'bathroom_cabinet', 'vase', 'light', 'coffee']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Sitting down and relaxing on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'scissors', 'drawing', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'shower', 'electric_shaver', 'food_kiwi', 'wall', 'bench', 'food_cheese', 'photoframe', 'newspaper', 'bed', 'food_food', 'standingmirror', 'television', 'curtain', 'doorjamb', 'shaving_cream', 'door', 'soap', 'clothes_shirt', 'scrabble', 'coffee']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to TV, grab remote control, press change channel button, repeat pressing change channel button util find the channel wanted, put back remote control,\n    \"actions\": ['[WALK] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[TOUCH] <button>', '[TURNTO] <television>', '[WATCH] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'blender', 'crayon', 'walllamp', 'ceilingfan', 'dough', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'mat', 'stovefan', 'wall', 'powersocket', 'oven', 'bench', 'cutting_board', 'food_rice', 'doorjamb', 'cd', 'door', 'light', 'vase', 'creditcard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I enter the bedroom and find the light. Then I turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'detergent', 'floor', 'ceiling', 'toaster', 'faucet', 'stereo', 'electric_shaver', 'window', 'closetdrawer', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'cutting_board', 'coffe_maker', 'television', 'drinking_glass', 'after_shave', 'shaving_cream', 'pencil', 'kitchen_counter', 'comb', 'cupboard', 'sink', 'bathroom_cabinet', 'vase', 'clothes_shirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <drinking_glass> <sink>', '[GRAB] <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'detergent', 'food_butter', 'dough', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'purse', 'faucet', 'stereo', 'picture', 'dustpan', 'bowl', 'dining_room', 'wall', 'oven', 'powersocket', 'photoframe', 'wine', 'tray', 'console', 'pencil', 'sink', 'bathroom_cabinet', 'dishwasher', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'food_salt', 'walllamp', 'detergent', 'trashcan', 'knifeblock', 'drawing', 'floor', 'stamp', 'table_cloth', 'plate', 'faucet', 'deck_of_cards', 'clothes_pants', 'wall', 'bench', 'powersocket', 'tape', 'photoframe', 'food_food', 'tray', 'pencil', 'kitchen_counter', 'cupboard', 'maindoor', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'coffee']\nTask:Bring dirty plate to sink\nresponse:\n{\n    \"thoughts\": Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[PUTBACK] <plate> <sink>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sponge', 'filing_cabinet', 'tea', 'chair', 'walllamp', 'bathroom', 'ceilingfan', 'character', 'computer', 'food_bacon', 'mop_bucket', 'balanceball', 'foundation', 'form', 'bowl', 'wall', 'pasta', 'basket_for_clothes', 'bed', 'check', 'doorjamb', 'diary', 'hairbrush', 'cupboard', 'dresser', 'laundry_detergent', 'food_steak', 'light', 'scrabble']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Listen to music", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'detergent', 'ceilingfan', 'hanger', 'bathroom_counter', 'phone', 'bedroom', 'bookshelf', 'keyboard', 'food_cake', 'clothes_jacket', 'button', 'laptop', 'tvstand', 'mat', 'dustpan', 'food_food', 'after_shave', 'towel_rack', 'pencil', 'kitchen_counter', 'laundry_detergent', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'table', 'vase', 'washing_machine', 'coffee']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'beer', 'chair', 'freezer', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'character', 'keyboard', 'bookmark', 'ceilinglamp', 'balanceball', 'window', 'mousepad', 'couch', 'dustpan', 'pillow', 'wall', 'food_cheese', 'coffe_maker', 'television', 'toothbrush_holder', 'mouse', 'pencil', 'maindoor', 'table', 'dishwasher', 'vase']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I go to the living room and sit in the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'filing_cabinet', 'beer', 'measuring_cup', 'chair', 'ground_coffee', 'knifeblock', 'longboard', 'floor', 'ceiling', 'bedroom', 'keyboard', 'computer', 'mail', 'toaster', 'picture', 'dustpan', 'wall', 'bench', 'cutting_board', 'tape', 'cup', 'coffe_maker', 'coffee_filter', 'water', 'after_shave', 'dresser', 'maindoor', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[GRAB] <ground_coffee>', '[PUTOBJBACK] <ground_coffee>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'walllamp', 'cat', 'drawing', 'needle', 'bathroom_counter', 'dvd_player', 'folder', 'desk', 'bedroom', 'computer', 'toaster', 'tvstand', 'balanceball', 'mousepad', 'dining_room', 'nightstand', 'oven', 'powersocket', 'food_food', 'pencil', 'maindoor', 'sink', 'bag', 'bathroom_cabinet', 'table', 'home_office', 'light', 'laser_pointer']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find the cat. Walk over to the cat. Touch the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'scissors', 'crayon', 'freezer', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'bookmark', 'toaster', 'ceilinglamp', 'stovefan', 'faucet', 'window', 'homework', 'couch', 'wall', 'child', 'bench', 'photoframe', 'newspaper', 'coffe_maker', 'razor', 'television', 'kitchen_counter', 'cupboard', 'table', 'coffee']\nTask:Let baby learn how to walk\nresponse:\n{\n    \"thoughts\": The baby tries to stand up and balances on the couch and there are no sharp edges so it's safe.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[TOUCH] <child>', '[GREET] <child>', '[FIND] <couch>', '[TURNTO] <couch>', '[POINTAT] <couch>', '[TURNTO] <child>', '[LOOKAT] <child>']\n}\n", "input": "\nScene_Objects:['dog', 'walllamp', 'ceilingfan', 'drawing', 'orchid', 'knife', 'longboard', 'floor', 'ceiling', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'oil', 'couch', 'candle', 'pillow', 'wall', 'keys', 'coffe_maker', 'television', 'coin', 'curtain', 'dresser', 'conditioner', 'door', 'sink', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "you go into the living room, sit on a chair/sofa, grab remote control, turn on tv, change channels, then watch somethign fun.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'dog', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bedroom', 'keyboard', 'computer', 'stovefan', 'faucet', 'oil', 'pillow', 'wall', 'wall_clock', 'oven', 'pasta', 'keys', 'check', 'tray', 'food_onion', 'kitchen_counter', 'cupboard', 'door', 'sink', 'bathtub', 'light']\nTask:Keep an eye on stove as something is cooking\nresponse:\n{\n    \"thoughts\": Watch whatever is cooking on the stove so it doesn't burn or pour. Turn off heat when done,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <oven>', '[FIND] <pot>', '[TURNTO] <pot>', '[LOOKAT] <pot>', '[FIND] <wall_clock>', '[TURNTO] <wall_clock>', '[LOOKAT] <wall_clock>', '[SWITCHOFF] <oven>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'knife', 'cpuscreen', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'powersocket', 'doorjamb', 'shaving_cream', 'mouse', 'bag', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "go to home office. turn the computer on. begin working", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'bed', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go the living room,  sit in the chair. Take the remote control and switch on the TV.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_pizza', 'juice', 'toilet', 'freezer', 'blow_dryer', 'walllamp', 'ceilingfan', 'knifeblock', 'bathroom_counter', 'alcohol', 'button', 'stereo', 'form', 'towel', 'oven', 'basket_for_clothes', 'photoframe', 'light_bulb', 'toilet_paper', 'toothbrush_holder', 'kitchen_counter', 'shoes', 'chessboard', 'vacuum_cleaner', 'clothes_scarf', 'home_office', 'dishwasher', 'soap', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and get the soap and wash my both hand using the soap and rinse my hands and use towel.", "actions": ["[WALK] <dining_room>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <soap>", "[FIND] <towel>", "[RINSE] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'basket_for_clothes', 'cup', 'television', 'food_rice', 'curtain', 'doorjamb', 'diary', 'shoes', 'door', 'spoon', 'bathroom_cabinet', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the sofa to watch tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'ceilingfan', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'mousepad', 'couch', 'food_kiwi', 'wall', 'bench', 'cup', 'check', 'television', 'curtain', 'napkin', 'table', 'soap', 'instrument_violin']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room. I sit on the couch. I pick up the remote control. I push the power button.  I push the guide button and look for my favorite show \"The Middle.\"  I click that channel and enjoy my show.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUSH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'orchid', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'stamp', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'dining_room', 'wall', 'toy', 'bed', 'food_food', 'toilet_paper', 'mouthwash', 'doorjamb', 'pencil', 'kitchen_counter', 'comb', 'dresser', 'laundry_detergent', 'napkin', 'home_office', 'bathroom_cabinet', 'dishwasher', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": \"I walk to entrance hall\" \"I walk to bedroom\" \"I find basket of clothes\" \"I lift basket of  clothes\" \"I walk to kitchen\" \"I find washing machine\" \"I open washing machine \" \"I put laundry in washing machine\" \"I add laundry detergent\" \" \"I close washing machine lid\" \"I switch washing machine on\",\n    \"actions\": ['[WALK] <home_office>', '[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[PUTBACK] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[TURNTO] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'ceilingfan', 'drawing', 'floor', 'bookshelf', 'ceilinglamp', 'shower', 'window', 'couch', 'food_orange', 'towel', 'dining_room', 'wall', 'powersocket', 'television', 'tray', 'wooden_spoon', 'rag', 'lighter', 'towel_rack', 'pencil', 'dresser', 'clothes_dress', 'hairdryer', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'clothes_shirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, find sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'bathroom', 'orchid', 'drawing', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'alcohol', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'bench', 'photoframe', 'food_food', 'television', 'curtain', 'doorjamb', 'door', 'sink', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find the television, switch on the television, find the sofa, walk to sofa, sit in sofa, watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'freezer', 'bathroom', 'ceilingfan', 'drawing', 'bills', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'toy', 'tape', 'bed', 'television', 'curtain', 'comb', 'hair', 'home_office', 'table', 'soap']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the living room and sit on the couch. Turn on television using remote control. Use remote control buttons to navigate to desired channel. Watch television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'walllamp', 'milk', 'floor', 'keyboard', 'table_cloth', 'stovefan', 'mat', 'tvstand', 'faucet', 'window', 'couch', 'clothes_hat', 'pillow', 'dining_room', 'toy', 'basket_for_clothes', 'coffe_maker', 'video_game_controller', 'tray', 'doorjamb', 'kettle', 'dresser', 'cupboard', 'clothes_dress', 'sink', 'table']\nTask:Pick up toys\nresponse:\n{\n    \"thoughts\": Bend down and pick up the toys off of the floor.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <dresser>', '[PUTIN] <toy> <dresser>', '[FIND] <toy>', '[WALK] <dresser>', '[GRAB] <toy>', '[PUTIN] <toy> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'detergent', 'ceilingfan', 'drawing', 'needle', 'knifeblock', 'floor', 'ceiling', 'phone', 'bedroom', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'bench', 'check', 'television', 'oven_mitts', 'curtain', 'conditioner', 'food_steak', 'home_office', 'table', 'vase', 'clothes_shirt']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk into the living room and sit on the sofa to watch tv", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'food_pizza', 'juice', 'board_game', 'dog', 'food_butter', 'orchid', 'drawing', 'bathroom_counter', 'bookshelf', 'stovefan', 'mat', 'laptop', 'tvstand', 'picture', 'window', 'couch', 'nightstand', 'wall_clock', 'toy', 'clothes_gloves', 'toilet_paper', 'television', 'ice', 'bathroom_cabinet', 'table', 'light', 'vase', 'instrument_violin']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk into kitchen, find plate, drop plate on table in living room, walk into kitchen, find cup, bring cup to table in living room, walk into kitchen find fork, bring fork to table in living room,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <plate>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <home_office>', '[DROP] <plate>', '[WALK] <dining_room>', '[WALK] <cup>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <home_office>', '[DROP] <cup>', '[WALK] <dining_room>', '[WALK] <fork>', '[FIND] <fork>', '[GRAB] <fork>', '[WALK] <home_office>', '[DROP] <fork>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'balanceball', 'mousepad', 'candle', 'wall', 'nightstand', 'newspaper', 'doorjamb', 'mouse', 'light', 'clothes_skirt']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Turn on computer, pull up browser, browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'door', 'home_office', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the remote. Pick up the remote. Find the sofa. Sit on the sofa. Look at the television. Aim the remote at the television. Press the power button. Flip the channels.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[PUSH] <remote_control>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'juice', 'ground_coffee', 'detergent', 'trashcan', 'hanger', 'floor', 'folder', 'desk', 'bedroom', 'bookshelf', 'mop_bucket', 'toaster', 'ceilinglamp', 'balanceball', 'microphone', 'clothes_hat', 'form', 'food_kiwi', 'nightstand', 'bench', 'photoframe', 'newspaper', 'kitchen_counter', 'hair', 'maindoor', 'door', 'food_steak', 'hairdryer']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the living room, walk to the lamp, switch on the lamp.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'desk', 'bookshelf', 'stovefan', 'ceilinglamp', 'faucet', 'stereo', 'shower', 'candle', 'wall', 'oven', 'cup', 'photoframe', 'water', 'tray', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'bathtub', 'bathroom_cabinet', 'light']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Walk to kitchen, walk to cupboard, open cupboard, grab cup, walk to sink, pour water in cup, drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <cup>', '[GRAB] <cup>', '[WALK] <sink>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <cup>', '[DRINK] <cup>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'food_butter', 'ceilingfan', 'trashcan', 'floor', 'ceiling', 'bedroom', 'mat', 'plate', 'balanceball', 'faucet', 'oil', 'clothes_pants', 'food_orange', 'wall', 'bench', 'tape', 'cup', 'bed', 'light_bulb', 'video_game_controller', 'razor', 'food_rice', 'diary', 'towel_rack', 'mouse', 'headset', 'dishrack', 'spoon']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room ,find radio ,switch on radio ,listen to music 4", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'walllamp', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'mat', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'basket_for_clothes', 'cup', 'television', 'food_rice', 'curtain', 'doorjamb', 'diary', 'shoes', 'door', 'spoon', 'bathroom_cabinet', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the sofa to watch tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['microwave', 'chair', 'nail_polish', 'knifeblock', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'shampoo', 'bookshelf', 'computer', 'ceilinglamp', 'food_kiwi', 'wall', 'bench', 'newspaper', 'bed', 'doorjamb', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'conditioner', 'door', 'hairdryer', 'light', 'table', 'tooth_paste', 'shoe_rack']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into living room. Find wall. Find light button. Switch on light button.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'microwave', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'cpuscreen', 'floor', 'desk', 'toaster', 'stovefan', 'plate', 'faucet', 'iron', 'dish_soap', 'wall', 'oven', 'coffe_maker', 'food_food', 'wooden_spoon', 'tray', 'kitchen_counter', 'dishrack', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['jelly', 'nail_polish', 'detergent', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'dvd_player', 'computer', 'ceilinglamp', 'plate', 'balanceball', 'faucet', 'window', 'stereo', 'couch', 'wall', 'oven', 'newspaper', 'television', 'tray', 'rag', 'mouse', 'hair', 'controller', 'laundry_detergent', 'napkin', 'home_office', 'table', 'creditcard']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music", "actions": ["[WALK] <home_office>", "[WALK] <dvd_player>", "[FIND] <dvd_player>", "[TURNTO] <dvd_player>", "[LOOKAT] <dvd_player>", "[SWITCHON] <dvd_player>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'food_salt', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'stereo', 'couch', 'wall', 'nightstand', 'bench', 'powersocket', 'tape', 'bed', 'coffe_maker', 'check', 'tray', 'curtain', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[PUSH] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['microwave', 'toilet', 'chair', 'band-aids', 'drawing', 'knifeblock', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'computer', 'bookmark', 'ceilinglamp', 'window', 'mousepad', 'couch', 'wall', 'nightstand', 'powersocket', 'television', 'tray', 'doorjamb', 'dresser', 'cupboard', 'door', 'light', 'table', 'dishwasher', 'board_game']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "My friend is calling and I want to pick up the phone and chat with him for a little while on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <phone>", "[GRAB] <phone>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>", "[TYPE] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'cat', 'orchid', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shower', 'mousepad', 'closetdrawer', 'wall', 'nightstand', 'cutting_board', 'coffee_table', 'novel', 'food_food', 'television', 'towel_rack', 'mouse', 'shoes', 'door', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I walk to the living room and pick up a book from the coffee table. I find my favorite chair.  I sit in the armchair and open the book.  I read.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[WALK] <coffee_table>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'envelope', 'computer', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'photoframe', 'check', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "I go to the home office to work on my project. I turn on the light and walk to my desk. I turn on the computer. Then I look through the documents. After I'm done, I turn off the computer.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <desk>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <check>", "[TURNTO] <check>", "[LOOKAT] <check>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'freezer', 'bathroom', 'orchid', 'phone', 'dvd_player', 'character', 'bedroom', 'ceilinglamp', 'balanceball', 'iron', 'window', 'couch', 'food_orange', 'wall', 'tape', 'cup', 'basket_for_clothes', 'check', 'wooden_spoon', 'console', 'mouse', 'kitchen_counter', 'hairbrush', 'food_steak', 'sink', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Wash hands with soap,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[WASH] <hands_both>', '[RINSE] <hands_both>']\n}\n", "input": "\nScene_Objects:['fork', 'water_glass', 'blender', 'food_salt', 'microwave', 'nail_polish', 'trashcan', 'drawing', 'hanger', 'longboard', 'floor', 'character', 'computer', 'food_bacon', 'mat', 'electric_shaver', 'mousepad', 'dining_room', 'bench', 'bed', 'food_food', 'video_game_controller', 'water', 'tray', 'doorjamb', 'mouse', 'kitchen_counter', 'bathroom_cabinet', 'table', 'dirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, hold drinking glass, lift glass, pour water in drinking glass", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <table>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[FIND] <water>", "[GRAB] <water>", "[WALK] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'beer', 'freezer', 'walllamp', 'bathroom', 'bathroom_counter', 'character', 'bedroom', 'desk', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'clothes_pants', 'towel', 'bench', 'basket_for_clothes', 'bed', 'coin', 'food_donut', 'curtain', 'towel_rack', 'mouse', 'cd', 'comb', 'laundry_detergent', 'bathtub', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['fork', 'toilet', 'chair', 'freezer', 'walllamp', 'bills', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'computer', 'toaster', 'mat', 'plate', 'faucet', 'couch', 'form', 'wall', 'wall_clock', 'television', 'food_onion', 'doorjamb', 'towel_rack', 'dresser', 'cupboard', 'door', 'light', 'table']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk in to the living room. I find the light switch. I turn on the lights.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'beer', 'tea', 'freezer', 'shampoo', 'character', 'desk', 'piano_bench', 'ceilinglamp', 'iron', 'picture', 'mousepad', 'couch', 'bowl', 'nightstand', 'wall_clock', 'toy', 'bed', 'coffe_maker', 'television', 'coin', 'after_shave', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'shoes', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": take remote control, push \"change channel\" button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[PUSH] <remote_control>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'tea', 'measuring_cup', 'toilet', 'walllamp', 'bathroom', 'knifeblock', 'orchid', 'floor', 'ceiling', 'folder', 'bookshelf', 'stovefan', 'ceilinglamp', 'faucet', 'mousepad', 'wall', 'nightstand', 'pasta', 'cup', 'food_food', 'television', 'water', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'light', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen, walk to cupboard, open cupboard, grab cup, walk to sink, pour water in cup, drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <sink>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'dog', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'phone', 'bookshelf', 'table_cloth', 'bookmark', 'toaster', 'mat', 'faucet', 'oven', 'cup', 'photoframe', 'check', 'television', 'towel_rack', 'kitchen_counter', 'chessboard', 'cupboard', 'door', 'sink', 'light', 'tooth_paste']\nTask:Pet dog\nresponse:\n{\n    \"thoughts\": Pet my dog in hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dog>', '[TURNTO] <dog>', '[LOOKAT] <dog>', '[TOUCH] <dog>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'measuring_cup', 'jelly', 'freezer', 'nail_polish', 'ceilingfan', 'drawing', 'orchid', 'longboard', 'floor', 'ceiling', 'bedroom', 'keyboard', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'deck_of_cards', 'mousepad', 'couch', 'clothes_pants', 'wall', 'television', 'curtain', 'kitchen_counter', 'chessboard', 'spoon', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk in living room. I find TV remote. I grab TV remote. I sit on sofa. I switch on TV. I switch channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'trashcan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'phone', 'character', 'bedroom', 'desk', 'bookshelf', 'computer', 'mat', 'tvstand', 'window', 'pillow', 'wall', 'powersocket', 'tray', 'towel_rack', 'mouse', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'bathtub', 'light', 'teeth', 'shoe_rack']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'beer', 'dog', 'freezer', 'walllamp', 'trashcan', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'clothes_jacket', 'toaster', 'ceilinglamp', 'plate', 'mousepad', 'microphone', 'food_orange', 'wall', 'wall_clock', 'bench', 'food_food', 'oven_mitts', 'coffee_filter', 'kitchen_counter', 'cd', 'cupboard', 'maindoor', 'home_office']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I pick up groceries. I open fridge. I put groceries in the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[DROP] <food_food>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'juice', 'freezer', 'walllamp', 'drawing', 'orchid', 'phone', 'ceiling', 'character', 'cleaning_solution', 'button', 'toaster', 'ceilinglamp', 'homework', 'oil', 'mousepad', 'pillow', 'dining_room', 'food_kiwi', 'pasta', 'oven', 'coffe_maker', 'food_food', 'television', 'rag', 'towel_rack', 'bathtub', 'instrument_guitar', 'vase']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Spray counter tops with all purpose cleaner. Using towel, wipe counter, removing all dirt and grime until clean.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[PUTOBJBACK] <cleaning_solution>', '[FIND] <rag>', '[GRAB] <rag>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['spectacles', 'glue', 'filing_cabinet', 'microwave', 'jelly', 'freezer', 'ceilingfan', 'orchid', 'food_egg', 'floor', 'phone', 'ceiling', 'desk', 'thread', 'mat', 'laptop', 'deck_of_cards', 'faucet', 'foundation', 'pillow', 'nightstand', 'powersocket', 'basket_for_clothes', 'food_food', 'lighter', 'sink', 'home_office', 'vase', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bedroom to find and grab basket of clothes. I then go to the bathroom and find the washing machine. I put clothes inside the washing machine and put washing powder. I switch on the washing machine.", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'cpuscreen', 'kitchen_cabinet', 'bathroom_counter', 'desk', 'bedroom', 'thread', 'faucet', 'shower', 'electric_shaver', 'microphone', 'towel', 'dining_room', 'wall', 'oven', 'cutting_board', 'photoframe', 'water', 'drinking_glass', 'curtain', 'food_onion', 'doorjamb', 'kettle', 'kitchen_counter', 'dresser', 'laundry_detergent', 'light', 'table', 'dirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": You walk to the kitchen and open the kitchen cabinet and grab a drinking glass. You find the water and pour the water into the drinking glass. You drink the water from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <drinking_glass>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'mop', 'freezer', 'blow_dryer', 'bathroom', 'detergent', 'milk', 'bathroom_counter', 'ceiling', 'character', 'mop_bucket', 'stovefan', 'shoe-shine_kit', 'balanceball', 'deck_of_cards', 'stereo', 'shower', 'mousepad', 'microphone', 'pillow', 'light_bulb', 'coffee_filter', 'doorjamb', 'towel_rack', 'pencil', 'cupboard', 'light', 'dishwasher']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "The task is browse and surf on the internet regarding a particular topic", "actions": ["[WALK] <bedroom>", "[WALK] <laptop>", "[FIND] <laptop>", "[SWITCHON] <laptop>", "[TURNTO] <laptop>", "[LOOKAT] <laptop>", "[WATCH] <laptop>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'window', 'mousepad', 'wall', 'powersocket', 'box', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'bathroom_cabinet', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to the Home Office.  Find the computer.  Turn on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['chair', 'food_carrot', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'character', 'bedroom', 'keyboard', 'piano_bench', 'bookmark', 'laptop', 'electric_shaver', 'food_kiwi', 'wall', 'powersocket', 'toothbrush_holder', 'curtain', 'rag', 'doorjamb', 'shaving_cream', 'cards', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'light', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the dining room and turn on the lights", "actions": ["[WALK] <dining_room>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I sit at my desk and work on the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'chair', 'detergent', 'milk', 'knife', 'drawing', 'cpuscreen', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'faucet', 'couch', 'wall', 'photoframe', 'bed', 'water', 'after_shave', 'mouse', 'controller', 'laundry_detergent', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Walk to sink and turn on the water. Grab the soap and wash the hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <water>", "[TURNTO] <water>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'microwave', 'freezer', 'blow_dryer', 'food_carrot', 'milk', 'drawing', 'floor', 'phone', 'keyboard', 'bookshelf', 'button', 'piano_bench', 'laptop', 'plate', 'oil', 'mousepad', 'wall', 'bench', 'photoframe', 'keys', 'food_food', 'check', 'tray', 'kitchen_counter', 'dishrack', 'hair', 'table', 'soap', 'teeth']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.,\n    \"actions\": ['[WALK] <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <fork>', '[GRAB] <fork>', '[PUTBACK] <fork> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>', '[GRAB] <plate>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'chair', 'bathroom', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'dvd_player', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mat', 'balanceball', 'window', 'mousepad', 'pillow', 'wall', 'nightstand', 'photoframe', 'newspaper', 'doorjamb', 'mouse', 'shoes', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go to home office, turn on the computer, open the internet and browse around the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'walllamp', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'floor_lamp', 'electric_shaver', 'wall', 'nightstand', 'powersocket', 'bed', 'doorjamb', 'shoes', 'hairbrush', 'door', 'light', 'table', 'soap']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to wall, find light switch, turn on lights,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[WALK] <bedroom>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'juice', 'food_salt', 'measuring_cup', 'ceilingfan', 'orchid', 'shampoo', 'bedroom', 'desk', 'table_cloth', 'toaster', 'tvstand', 'stereo', 'plate', 'bench', 'basket_for_clothes', 'coffee_filter', 'ice', 'tray', 'pencil', 'headset', 'cupboard', 'maindoor', 'napkin', 'bag', 'light', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'walllamp', 'nail_polish', 'trashcan', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'toaster', 'stovefan', 'faucet', 'towel', 'pillow', 'wall', 'wall_clock', 'bed', 'check', 'television', 'doorjamb', 'kitchen_counter', 'dresser', 'controller', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Cleaning hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'spectacles', 'food_bread', 'microwave', 'walllamp', 'ground_coffee', 'detergent', 'longboard', 'floor', 'character', 'desk', 'shower', 'picture', 'balanceball', 'food_peanut_butter', 'pillow', 'dining_room', 'pasta', 'powersocket', 'basket_for_clothes', 'cup', 'food_food', 'razor', 'after_shave', 'mouse', 'paper_towel', 'cupboard', 'maindoor', 'board_game', 'hands_both']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walking towards Toilet room, And dispose of human waste. Cleaning your hands with tissue paper or water.", "actions": ["[WALK] <bathroom>", "[WALK] <paper_towel>", "[FIND] <paper_towel>", "[TURNTO] <paper_towel>", "[LOOKAT] <paper_towel>", "[GRAB] <paper_towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'curtain', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": Find home office. Walk to home office. Find computer. Find chair at desk. Sit in chair. Switch on computer. Work on computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'toilet', 'chair', 'dog', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'mat', 'balanceball', 'mousepad', 'bowl', 'wall', 'nightstand', 'food_food', 'check', 'television', 'doorjamb', 'mouse', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "I go to the home office to find the computer. I turn switch computer on , i open the email. i write the email and  send it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'freezer', 'dough', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'toaster', 'mat', 'pillow', 'dining_room', 'wall', 'oven', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'curtain', 'centerpiece', 'doorjamb', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'bag', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['sheets', 'dry_pasta', 'pot', 'food_salt', 'freezer', 'food_carrot', 'drawing', 'food_egg', 'floor', 'shampoo', 'ceiling', 'toaster', 'ceilinglamp', 'tvstand', 'shower', 'window', 'mousepad', 'couch', 'wall', 'food_food', 'television', 'lighter', 'dresser', 'cupboard', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'vase', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I sit on the sofa. I put my legs on the sofa and recline.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'window', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'photoframe', 'television', 'coin', 'doorjamb', 'mouse', 'chessboard', 'bathtub', 'table']\nTask:Write  school paper\nresponse:\n{\n    \"thoughts\": Writing  and drafting a paper for  class work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'detergent', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'dvd_player', 'folder', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'tvstand', 'shower', 'homework', 'oil', 'dustpan', 'dining_room', 'wall', 'powersocket', 'bench', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'door', 'home_office', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk into the bedroom and walk to the light switch. Flip the switch.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_salt', 'freezer', 'orchid', 'wallshelf', 'clothes_underwear', 'cpuscreen', 'phone', 'character', 'bedroom', 'keyboard', 'toaster', 'mat', 'plate', 'picture', 'foundation', 'dish_soap', 'clothes_socks', 'pillow', 'powersocket', 'cutting_board', 'basket_for_clothes', 'bed', 'television', 'curtain', 'diary', 'cupboard', 'sink', 'spoon', 'home_office']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Pick up dish.  Put in soapy water.  Wash dish.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dish_soap>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <dish_soap> <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WIPE] <plate>']\n}\n", "input": "\nScene_Objects:['sheets', 'jelly', 'detergent', 'orchid', 'knifeblock', 'drawing', 'hanger', 'longboard', 'bedroom', 'desk', 'envelope', 'computer', 'button', 'window', 'foundation', 'mousepad', 'couch', 'pasta', 'bench', 'powersocket', 'cup', 'clothes_gloves', 'video_game_controller', 'coffee_filter', 'centerpiece', 'towel_rack', 'hairbrush', 'cupboard', 'vase', 'electrical_outlet']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "walk to living room ,find electrical outlet , switch on eletrical outlet", "actions": ["[WALK] <home_office>", "[WALK] <electrical_outlet>", "[FIND] <electrical_outlet>", "[SWITCHON] <electrical_outlet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'remote_control', 'computer', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'powersocket', 'oven', 'coffe_maker', 'television', 'coffee_filter', 'food_rice', 'tray', 'curtain', 'pencil', 'kitchen_counter', 'cupboard', 'sink', 'napkin', 'table', 'dirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I like to set on my sofa.Turn on the tv. And watch Perry Mason. Monday's through Friday's.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_bread', 'toilet', 'chair', 'freezer', 'bathroom', 'food_carrot', 'hanger', 'keyboard', 'envelope', 'mail', 'tvstand', 'clothes_pants', 'bowl', 'dining_room', 'oven', 'powersocket', 'bed', 'clothes_gloves', 'check', 'wine', 'curtain', 'food_onion', 'kettle', 'dresser', 'vase', 'home_office', 'light', 'clothes_skirt', 'dishwasher']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I hear the phone ringing in the living room. I walk to the living room and go to the phone. I pick it up and answer the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>", "[SWITCHON] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'dustpan', 'pillow', 'wall', 'wall_clock', 'powersocket', 'television', 'tray', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Browsing the internet for information.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'microwave', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'orchid', 'floor', 'ceiling', 'character', 'thread', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'clothes_hat', 'food_orange', 'wall', 'oven', 'box', 'food_cheese', 'food_food', 'television', 'food_rice', 'after_shave', 'dresser', 'clothes_scarf', 'light', 'clothes_skirt', 'tooth_paste']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I take the groceries out of the grocery bags, throw away grocery bags, categorize the food so that I know where it goes in the fridge, open the fridge, and place each item in the fridge before closing the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'measuring_cup', 'toilet', 'drawing', 'needle', 'hanger', 'floor', 'computer', 'mat', 'mousepad', 'wall', 'nightstand', 'powersocket', 'cutting_board', 'bed', 'novel', 'food_food', 'video_game_controller', 'coffe_maker', 'tray', 'food_onion', 'doorjamb', 'lighter', 'cd', 'clothes_dress', 'bathroom_cabinet', 'light', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I would go lie down in my bed and open the book and start reading.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'trashcan', 'food_carrot', 'drawing', 'knifeblock', 'hanger', 'floor', 'tea_bag', 'ceilinglamp', 'tvstand', 'shower', 'window', 'homework', 'couch', 'dustpan', 'food_orange', 'dining_room', 'wall', 'nightstand', 'photoframe', 'clothes_gloves', 'coffe_maker', 'check', 'television', 'mouse', 'pencil', 'table', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room, turn to the sofa. Sit on the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'freezer', 'trashcan', 'food_carrot', 'orchid', 'floor', 'ceiling', 'thread', 'mail', 'button', 'food_bacon', 'bookmark', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'wall', 'bench', 'cup', 'food_food', 'razor', 'television', 'food_donut', 'towel_rack', 'console', 'kitchen_counter', 'hairbrush', 'cupboard', 'sink', 'vase']\nTask:Added meat to freezer\nresponse:\n{\n    \"thoughts\": After shopping for groceries, i returned home, i walked to the dining room and placed meat items in freezer located in the dining room. I opened up the bags from the grocery store, removed all the meat items, and after opening the freezer door, i placed the meat items into the freezer, and closed the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTBACK] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toothbrush', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'bookshelf', 'keyboard', 'computer', 'toaster', 'mat', 'tvstand', 'towel', 'wall', 'powersocket', 'basket_for_clothes', 'coffe_maker', 'coffee_filter', 'food_rice', 'after_shave', 'doorjamb', 'cards', 'door', 'napkin', 'light', 'vase']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the bedroom and use the button on the wall, next to the door to turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'chair', 'drawing', 'food_egg', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'bookshelf', 'mat', 'laptop', 'faucet', 'wall', 'box', 'photoframe', 'coffe_maker', 'colander', 'curtain', 'pencil', 'headset', 'dishrack', 'pajamas', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'clothes_shirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk in to the bathroom. I find the soap. I find the faucet. I turn on the faucet. I wash my hands. I put the soap back. I rinse my hands. I turn the faucet off.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'chair', 'freezer', 'ground_coffee', 'walllamp', 'food_carrot', 'bathroom_counter', 'floor', 'ceiling', 'toaster', 'ceilinglamp', 'clothes_socks', 'clothes_pants', 'wall', 'pasta', 'bench', 'powersocket', 'coffe_maker', 'check', 'food_food', 'coffee_filter', 'tray', 'doorjamb', 'kettle', 'kitchen_counter', 'napkin', 'bathroom_cabinet', 'dishwasher', 'vase']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Make coffee in the kitchen.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[PLUGIN] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'picture', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'check', 'television', 'ice', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'table']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": walk to living room.grab remote control. look at television. switch on television. lift remote control. push remote control. watch television. switch off television. put back remote control.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[SWITCHON] <television>', '[PUSH] <remote_control>', '[WATCH] <television>', '[SWITCHOFF] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['juice', 'crayon', 'walllamp', 'trashcan', 'dough', 'floor', 'phone', 'bedroom', 'tea_bag', 'toaster', 'mat', 'plate', 'ceilinglamp', 'dish_soap', 'wall', 'pasta', 'oven', 'cutting_board', 'food_cheese', 'newspaper', 'coffe_maker', 'check', 'food_food', 'tray', 'towel_rack', 'pencil', 'dresser', 'conditioner', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'nail_polish', 'detergent', 'orchid', 'drawing', 'hanger', 'floor', 'phone', 'desk', 'button', 'ceilinglamp', 'stereo', 'faucet', 'mousepad', 'dining_room', 'nightstand', 'wall_clock', 'oven', 'cup', 'photoframe', 'bed', 'food_food', 'tray', 'doorjamb', 'kitchen_counter', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I go into the living room and turn on the radio.  Then, I select the station I want, which is usually a preset.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[FIND] <button>', '[PUSH] <button>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['tea', 'walllamp', 'detergent', 'cat', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'remote_control', 'mail', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'clothes_socks', 'couch', 'wall', 'bench', 'clothes_gloves', 'coffe_maker', 'television', 'curtain', 'coffee_pot', 'home_office', 'table', 'coffee']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk to living room, look at the television, switch on the television, grab the remote control, press the channel button in remote control, now the tv channel is changed", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[GRAB] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'trashcan', 'cpuscreen', 'clothes_underwear', 'ceiling', 'folder', 'desk', 'computer', 'stovefan', 'shower', 'mousepad', 'candle', 'clothes_hat', 'dining_room', 'newspaper', 'photoframe', 'food_food', 'video_game_controller', 'oven_mitts', 'curtain', 'pencil', 'kitchen_counter', 'dishrack', 'laundry_detergent', 'door', 'light', 'dishwasher', 'soap', 'tooth_paste', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'jelly', 'ceilingfan', 'food_carrot', 'drawing', 'knife', 'clothes_underwear', 'floor', 'phone', 'ceilinglamp', 'tvstand', 'stereo', 'window', 'mousepad', 'couch', 'food_jam', 'wall', 'nightstand', 'cup', 'television', 'ice', 'mouse', 'kitchen_counter', 'home_office', 'table', 'soap', 'board_game', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "sitting down to listen to music on the radio", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'microwave', 'band-aids', 'drawing', 'cpuscreen', 'hanger', 'floor', 'computer', 'ceilinglamp', 'balanceball', 'window', 'oil', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'coffe_maker', 'toilet_paper', 'television', 'kitchen_counter', 'chessboard', 'home_office', 'table', 'bathroom_cabinet', 'vase', 'light', 'coffee']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Sitting down and relaxing on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'computer', 'toaster', 'ceilinglamp', 'window', 'food_noodles', 'couch', 'microphone', 'pillow', 'wall', 'newspaper', 'novel', 'coffe_maker', 'television', 'coin', 'toothbrush_holder', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'bathroom_cabinet', 'table', 'vase', 'laser_pointer', 'creditcard']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Sit on couch. Grab book. Open book. Read book.", "actions": ["[WALK] <couch>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'desk', 'keyboard', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'wall', 'pasta', 'oven', 'powersocket', 'box', 'water', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'sauce_pan', 'cupboard', 'door', 'light', 'dishwasher', 'tooth_paste', 'teeth']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Cook some pasta on the stove,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pasta>', '[GRAB] <pasta>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <sauce_pan>', '[POUR] <water> <sauce_pan>', '[FIND] <oven>', '[SWITCHON] <oven>', '[PUTBACK] <pasta> <sauce_pan>', '[FIND] <sauce_pan>', '[GRAB] <sauce_pan>', '[PUTBACK] <sauce_pan> <oven>']\n}\n", "input": "\nScene_Objects:['sponge', 'microwave', 'chair', 'walllamp', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'computer', 'stovefan', 'ceilinglamp', 'toaster', 'mousepad', 'pillow', 'wall', 'cutting_board', 'box', 'toilet_paper', 'doorjamb', 'centerpiece', 'towel_rack', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'door', 'light', 'dishwasher', 'shoe_rack']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to living room. Go near telephone. take the receiver and then talk to other person on that side.", "actions": ["[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'microwave', 'chair', 'walllamp', 'bathroom', 'detergent', 'bathroom_counter', 'button', 'toaster', 'brush', 'stereo', 'shower', 'faucet', 'food_orange', 'pillow', 'oven', 'photoframe', 'check', 'television', 'tray', 'kettle', 'mouse', 'cupboard', 'conditioner', 'door', 'home_office', 'table', 'dishwasher', 'clothes_skirt', 'soap']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk into living room, look for radio, grab on button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[TURNTO] <stereo>', '[LOOKAT] <stereo>', '[FIND] <button>', '[GRAB] <button>']\n}\n", "input": "\nScene_Objects:['scissors', 'mop', 'ground_coffee', 'ceilingfan', 'trashcan', 'toothbrush', 'drawing', 'needle', 'bookshelf', 'keyboard', 'laptop', 'balanceball', 'stereo', 'mousepad', 'wall_clock', 'bench', 'coffe_maker', 'food_food', 'toothbrush_holder', 'television', 'curtain', 'doorjamb', 'kettle', 'controller', 'napkin', 'home_office', 'light', 'vase', 'soap', 'tooth_paste']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "walk to bathroom, look at the toothbrush holder, take toothbrush from the toothbrush holder, find the toothpaste, squeeze the toothpaste into toothbrush, now brush teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush_holder>", "[FIND] <toothbrush_holder>", "[TURNTO] <toothbrush_holder>", "[LOOKAT] <toothbrush_holder>", "[FIND] <toothbrush>", "[TURNTO] <toothbrush>", "[LOOKAT] <toothbrush>", "[FIND] <tooth_paste>", "[SQUEEZE] <tooth_paste>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'dog', 'orchid', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'bookmark', 'mat', 'ceilinglamp', 'tvstand', 'laptop', 'mousepad', 'wall', 'nightstand', 'oven', 'photoframe', 'novel', 'television', 'lighter', 'towel_rack', 'mouse', 'dresser', 'door']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Reading a book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <bookmark>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['ground_coffee', 'nail_polish', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'remote_control', 'computer', 'button', 'mail', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'tape', 'television', 'curtain', 'food_onion', 'cupboard', 'table', 'shoe_rack']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the living room, sit on the couch, find the remote, switch on the TV and watch", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'milk', 'drawing', 'orchid', 'needle', 'hanger', 'bathroom_counter', 'floor', 'phone', 'computer', 'bookmark', 'toaster', 'mat', 'tvstand', 'electric_shaver', 'clothes_socks', 'closetdrawer', 'couch', 'nightstand', 'wall', 'bed', 'keys', 'photoframe', 'tray', 'centerpiece', 'dresser', 'bag', 'soap', 'clothes_shirt', 'teeth']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Climb into bed, then read the novel,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <bed>', '[LIE] <bed>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'freezer', 'bathroom', 'orchid', 'phone', 'dvd_player', 'character', 'bedroom', 'ceilinglamp', 'balanceball', 'iron', 'window', 'couch', 'food_orange', 'wall', 'tape', 'cup', 'basket_for_clothes', 'check', 'wooden_spoon', 'console', 'mouse', 'kitchen_counter', 'hairbrush', 'food_steak', 'sink', 'home_office', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Wash hands with soap", "actions": ["[WALK] <bathroom>", "[WALK] <soap>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'orchid', 'bills', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'shoe-shine_kit', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'bed', 'television', 'doorjamb', 'mouse', 'vacuum_cleaner', 'comb', 'door', 'home_office', 'light', 'table']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I go to the living room, where my computer is. I locate it, switch it on and type desired URL adress in the web browser.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'milk', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'homework', 'towel', 'dining_room', 'wall', 'nightstand', 'oven', 'food_food', 'tray', 'console', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'bag', 'napkin', 'soap']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to kitchen counter, grab a grocery, walk to fridge, open fridge, put the grocery into fridge, and close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'toilet', 'chair', 'detergent', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'character', 'bookshelf', 'purse', 'envelope', 'toaster', 'tvstand', 'shower', 'window', 'pillow', 'tape', 'cup', 'coffe_maker', 'check', 'oven_mitts', 'curtain', 'rag', 'lighter', 'mouse', 'pencil', 'hairbrush', 'table', 'vase']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find the cat, go near the cat and pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['fork', 'juice', 'tea', 'walllamp', 'toothbrush', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'character', 'button', 'table_cloth', 'brush', 'mat', 'faucet', 'food_jam', 'wall', 'powersocket', 'photoframe', 'mouse', 'vacuum_cleaner', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I walk in to the bathroom. I turn on the faucet. I find the toothbrush and toothpaste. I grab the toothbrush and toothpaste. I rinse off the toothbrush. I add toothpaste to the toothbrush. I put back the toothpaste. I wash my teeth with the toothbrush. I rinse off the toothbrush. I put the toothbrush back. I rinse my mouth out. I turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[RINSE] <toothbrush>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>", "[PUTOBJBACK] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>", "[RINSE] <toothbrush>", "[PUTOBJBACK] <toothbrush>", "[RINSE] <teeth>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'filing_cabinet', 'tablelamp', 'mop', 'microwave', 'toilet', 'band-aids', 'bathroom', 'trashcan', 'wallshelf', 'character', 'computer', 'mat', 'faucet', 'couch', 'dining_room', 'wall', 'newspaper', 'oven_mitts', 'tray', 'curtain', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'sink', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Shut front door\nresponse:\n{\n    \"thoughts\": Close the storm door. Lock the storm door.,\n    \"actions\": ['[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'crayon', 'freezer', 'walllamp', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'shampoo', 'desk', 'mail', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'stovefan', 'mat', 'electric_shaver', 'mousepad', 'clothes_pants', 'wall', 'nightstand', 'pasta', 'basket_for_clothes', 'food_food', 'diary', 'towel_rack', 'dresser', 'clothes_dress']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "in my home i walk to the kitchen with some groceries and locate the fridge,then open the fridge put the groceries into the fridge and closed.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'jelly', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'desk', 'keyboard', 'computer', 'toaster', 'mat', 'laptop', 'stovefan', 'tvstand', 'window', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'powersocket', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'table']\nTask:Use laptop\nresponse:\n{\n    \"thoughts\": Sit down on a couch and perform tasks such as navigating the internet on a laptop,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <laptop>', '[SWITCHON] <laptop>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <keyboard>', '[LOOKAT] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <laptop>', '[LOOKAT] <laptop>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_salt', 'beer', 'walllamp', 'knifeblock', 'cpuscreen', 'floor', 'bookshelf', 'mail', 'computer', 'button', 'toaster', 'plate', 'shower', 'dish_soap', 'mousepad', 'pillow', 'wall', 'oven', 'photoframe', 'clothes_gloves', 'coffe_maker', 'tray', 'food_onion', 'doorjamb', 'dresser', 'hairdryer', 'sink', 'dishwasher', 'clothes_shirt']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Put dishes in the dishwasher and start the dishwasher.", "actions": ["[RUN] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[WALK] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[WALK] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'freezer', 'blow_dryer', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'bedroom', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'shoe-shine_kit', 'couch', 'microphone', 'food_jam', 'wall', 'coffe_maker', 'check', 'television', 'curtain', 'sink', 'napkin', 'bathroom_cabinet', 'table', 'colander', 'tooth_paste']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find the television. Walk to the television. Switch on the television. Find the sofa. Walk to the sofa. Sit on the sofa. Find the remote control. Grab the remote control. Look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[WALK] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'sheets', 'mop', 'microwave', 'chair', 'freezer', 'blow_dryer', 'nail_polish', 'orchid', 'knife', 'floor', 'keyboard', 'stovefan', 'stereo', 'foundation', 'couch', 'soap', 'food_orange', 'pillow', 'wall', 'powersocket', 'check', 'toilet_paper', 'coffe_maker', 'curtain', 'pencil', 'shoes', 'vacuum_cleaner', 'colander', 'shoe_rack']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Petting the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'tablelamp', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'window', 'oil', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'nightstand', 'wall', 'oven', 'bed', 'coffe_maker', 'television', 'curtain', 'dresser', 'door', 'instrument_guitar', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go to the living room, switch on the tv, take the remote and sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['blender', 'measuring_cup', 'microwave', 'freezer', 'trashcan', 'food_butter', 'orchid', 'longboard', 'floor', 'desk', 'keyboard', 'mat', 'plate', 'ceilinglamp', 'homework', 'couch', 'candle', 'clothes_pants', 'bench', 'bed', 'check', 'mouthwash', 'television', 'food_onion', 'controller', 'maindoor', 'sink', 'spoon', 'napkin', 'bathroom_cabinet']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch, flip the switch to the on position", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'trashcan', 'toothbrush', 'drawing', 'orchid', 'needle', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'bookshelf', 'remote_control', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'mat', 'wall', 'powersocket', 'television', 'food_rice', 'doorjamb', 'pencil', 'chessboard', 'door', 'hairdryer', 'bathroom_cabinet', 'table', 'scrabble']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go to the livingroom. The TV is on. I pick up the remote and press the next channel button. I look at the television, and see that the channel has changed.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'filing_cabinet', 'tea', 'toilet', 'ceilingfan', 'drawing', 'orchid', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'clothes_jacket', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'couch', 'form', 'dining_room', 'wall', 'nightstand', 'oven', 'coffe_maker', 'television', 'oven_mitts', 'curtain', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to the room, in which the TV is located. Turn on TV pressing the power button for the TV. Watch TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'ground_coffee', 'walllamp', 'food_butter', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'thread', 'stovefan', 'toaster', 'tvstand', 'faucet', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'coffe_maker', 'food_food', 'oven_mitts', 'coffee_filter', 'water', 'wine', 'wooden_spoon', 'controller', 'bathroom_cabinet', 'home_office', 'dishwasher']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to kitchen, open coffee maker, find coffee filter, grab coffee filter, put coffee filter in coffee maker, find ground coffee, put ground coffee in coffee maker, pour water in coffee maker, close coffee maker and switch on coffee maker.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'jelly', 'food_egg', 'hanger', 'desk', 'bedroom', 'food_bacon', 'toaster', 'mat', 'balanceball', 'window', 'food_peanut_butter', 'clothes_hat', 'tape', 'basket_for_clothes', 'photoframe', 'clothes_gloves', 'newspaper', 'coffe_maker', 'food_food', 'coin', 'after_shave', 'kitchen_counter', 'laundry_detergent', 'napkin', 'light', 'table', 'washing_machine', 'teeth']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to bedroom and grab dirty pants and take to the washing machine in entrance hall. Put clothes in washing machine and then put in laundry detergent.", "actions": ["[WALK] <bedroom>", "[WALK] <clothes_pants>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'juice', 'microwave', 'freezer', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'tea_bag', 'keyboard', 'computer', 'button', 'stovefan', 'mat', 'ceilinglamp', 'homework', 'closetdrawer', 'couch', 'wall', 'oven', 'cutting_board', 'toilet_paper', 'television', 'curtain', 'diary', 'pencil', 'shoes', 'dresser', 'table']\nTask:Put out flowers\nresponse:\n{\n    \"thoughts\": Place a flower centrepiece on the entrance hall table.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <centerpiece>', '[FIND] <centerpiece>', '[GRAB] <centerpiece>', '[WALK] <table>', '[PUTBACK] <centerpiece> <table>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'blender', 'microwave', 'toilet', 'dog', 'freezer', 'drawing', 'knifeblock', 'longboard', 'floor', 'folder', 'computer', 'ceilinglamp', 'window', 'couch', 'wall', 'nightstand', 'bench', 'cup', 'bed', 'clothes_gloves', 'novel', 'television', 'curtain', 'towel_rack', 'cupboard', 'sink', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I'm tired and want to relax on the sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'bathroom', 'food_butter', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'mop_bucket', 'toaster', 'shower', 'closetdrawer', 'pillow', 'dining_room', 'wall', 'nightstand', 'powersocket', 'toy', 'bed', 'razor', 'television', 'after_shave', 'doorjamb', 'dresser', 'conditioner', 'door', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into the bedroom and walk to the light switch. Flip the switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'pot', 'tablelamp', 'measuring_cup', 'freezer', 'walllamp', 'ceilingfan', 'knifeblock', 'drawing', 'longboard', 'floor', 'ceiling', 'keyboard', 'button', 'clothes_jacket', 'tvstand', 'couch', 'form', 'dining_room', 'bowl', 'nightstand', 'light_bulb', 'check', 'curtain', 'mouse', 'hair', 'cupboard', 'hairdryer', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find water glass. Grab water glass. Find water. Pour water in water glass. Drink water from water glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'filing_cabinet', 'walllamp', 'orchid', 'bathroom_counter', 'floor', 'phone', 'character', 'desk', 'computer', 'brush', 'toaster', 'mat', 'shower', 'faucet', 'electric_shaver', 'mousepad', 'dining_room', 'nightstand', 'oven', 'toy', 'cup', 'food_onion', 'mouse', 'dresser', 'cupboard', 'door', 'bathroom_cabinet', 'dishwasher']\nTask:Put toys away\nresponse:\n{\n    \"thoughts\": Pick up all the toys on the floor and put them in their correct storage bin or shelf.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <toy>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[FIND] <toy>', '[GRAB] <toy>', '[WALK] <filing_cabinet>', '[PUTBACK] <toy> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['sheets', 'water_glass', 'scissors', 'crayon', 'freezer', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'keyboard', 'thread', 'table_cloth', 'stovefan', 'ceilinglamp', 'shower', 'foundation', 'wall', 'nightstand', 'toy', 'water', 'ice', 'tray', 'wooden_spoon', 'vacuum_cleaner', 'door', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "go to kitchen, find water glass and grab it, find fridge and open grab water and pour into it, then drink.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[FIND] <freezer>", "[OPEN] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[CLOSE] <freezer>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'tablelamp', 'dog', 'detergent', 'needle', 'hanger', 'floor', 'ceiling', 'envelope', 'table_cloth', 'ceilinglamp', 'broom', 'wall', 'cutting_board', 'photoframe', 'bed', 'coffe_maker', 'check', 'coffee_filter', 'water', 'kitchen_counter', 'comb', 'dresser', 'cupboard', 'controller', 'table', 'dishwasher', 'clothes_skirt', 'coffee']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": I find the coffee machine, fill it with water and coffee then press the button to make the coffee,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[WALK] <coffee>', '[FIND] <coffee>', '[GRAB] <coffee>', '[POUR] <coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['beer', 'crayon', 'ceilingfan', 'drawing', 'needle', 'floor', 'ceiling', 'shampoo', 'keyboard', 'remote_control', 'thread', 'food_bacon', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'photoframe', 'bed', 'television', 'coffee_filter', 'curtain', 'mouse', 'table', 'light', 'laser_pointer']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I go into the living room. I find the remote. I turn on the tv. I sit on the couch and watch tv.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'beer', 'toilet', 'chair', 'ceilingfan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'folder', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'microphone', 'wall', 'bench', 'light_bulb', 'television', 'curtain', 'food_steak', 'sink', 'napkin', 'home_office', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into my living room, find the remote and sit on my sofa. I look at the remote and switch on the television. I look at the Television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['scissors', 'cat', 'drawing', 'hanger', 'longboard', 'floor', 'phone', 'character', 'desk', 'purse', 'stovefan', 'ceilinglamp', 'iron', 'stereo', 'window', 'oil', 'couch', 'wall', 'bench', 'toy', 'food_food', 'television', 'food_rice', 'towel_rack', 'console', 'hairbrush', 'napkin', 'table', 'vase', 'dirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I would walk into the living room. Then sit on the couch and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'freezer', 'ground_coffee', 'food_carrot', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'brush', 'ceilinglamp', 'mat', 'picture', 'clothes_socks', 'couch', 'microphone', 'food_orange', 'wall', 'powersocket', 'oven', 'tape', 'cup', 'keys', 'novel', 'food_food', 'check', 'mouthwash', 'curtain', 'pajamas', 'sink']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": take every item from groceries bag and put them in the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['food_bread', 'jelly', 'drawing', 'knifeblock', 'floor', 'computer', 'food_bacon', 'mail', 'ceilinglamp', 'shoe-shine_kit', 'tvstand', 'shower', 'window', 'couch', 'food_kiwi', 'wall', 'box', 'food_food', 'television', 'coffee_filter', 'after_shave', 'doorjamb', 'kettle', 'mouse', 'hair', 'cupboard', 'conditioner', 'sink', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "walk to living room, walk to sofa, sit in sofa", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'freezer', 'trashcan', 'milk', 'knifeblock', 'orchid', 'clothes_underwear', 'bookshelf', 'purse', 'thread', 'computer', 'ceilinglamp', 'tvstand', 'iron', 'couch', 'microphone', 'wall', 'cutting_board', 'basket_for_clothes', 'keys', 'toothbrush_holder', 'oven_mitts', 'curtain', 'doorjamb', 'diary', 'chessboard', 'home_office', 'bathroom_cabinet', 'clothes_skirt', 'light']\nTask:Hang up car keys\nresponse:\n{\n    \"thoughts\": Open the front door. Enter the entrance hall. Take out car keys from purse. Hang car keys up on hook on the wall.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <purse>', '[FIND] <purse>', '[OPEN] <purse>', '[FIND] <keys>', '[GRAB] <keys>', '[CLOSE] <purse>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'cd_player', 'toilet', 'freezer', 'trashcan', 'food_carrot', 'milk', 'music_stand', 'clothes_underwear', 'longboard', 'desk', 'bedroom', 'tea_bag', 'stovefan', 'ceilinglamp', 'faucet', 'electric_shaver', 'mousepad', 'food_jam', 'wall', 'nightstand', 'toy', 'bed', 'food_food', 'kitchen_counter', 'door', 'sink', 'dishwasher', 'coffee']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to the living and turn to the music stand. Switch on the CD player to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <music_stand>", "[FIND] <music_stand>", "[TURNTO] <music_stand>", "[FIND] <cd_player>", "[SWITCHON] <cd_player>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'food_pizza', 'filing_cabinet', 'tablelamp', 'freezer', 'nail_polish', 'cat', 'food_carrot', 'drawing', 'orchid', 'hanger', 'kitchen_cabinet', 'floor', 'ceiling', 'stamp', 'ceilinglamp', 'plate', 'candle', 'dining_room', 'wall', 'bench', 'food_vegetable', 'newspaper', 'food_food', 'television', 'after_shave', 'cards', 'cd', 'clothes_dress', 'table']\nTask:Fix snack\nresponse:\n{\n    \"thoughts\": Open fridge. Take out carrots and other vegetables. Put vegetables on a plate. Put away carrots and celery. Close fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[CLOSE] <kitchen_cabinet>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_carrot>', '[GRAB] <food_carrot>', '[PUTBACK] <food_carrot> <plate>', '[FIND] <food_vegetable>', '[GRAB] <food_vegetable>', '[PUTBACK] <food_vegetable> <plate>', '[CLOSE] <freezer>', '[WALK] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['glue', 'food_pizza', 'filing_cabinet', 'toilet', 'walllamp', 'detergent', 'orchid', 'food_egg', 'floor', 'phone', 'ceiling', 'character', 'keyboard', 'alcohol', 'toaster', 'tvstand', 'faucet', 'balanceball', 'picture', 'wall', 'oven', 'tape', 'food_cheese', 'photoframe', 'newspaper', 'coffe_maker', 'tray', 'towel_rack', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the sink. Take dishes out of sink. Open dishwasher and place dishes in dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'chair', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'bills', 'cpuscreen', 'bathroom_counter', 'bedroom', 'computer', 'mail', 'bookmark', 'shower', 'couch', 'towel', 'pillow', 'wall_clock', 'oven', 'powersocket', 'bench', 'bed', 'photoframe', 'television', 'mirror', 'curtain', 'towel_rack', 'kitchen_counter', 'home_office', 'dirt']\nTask:Look at mirror\nresponse:\n{\n    \"thoughts\": I look at my reflection to check my cleanliness.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <mirror>', '[FIND] <mirror>', '[TURNTO] <mirror>', '[LOOKAT] <mirror>']\n}\n", "input": "\nScene_Objects:['beer', 'freezer', 'band-aids', 'ceilingfan', 'food_carrot', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'food_chicken', 'character', 'ceilinglamp', 'faucet', 'electric_shaver', 'couch', 'microphone', 'wall', 'bench', 'food_food', 'toothbrush_holder', 'television', 'lighter', 'mouse', 'kitchen_counter', 'dresser', 'controller', 'table', 'clothes_shirt', 'teeth']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Put the groceries on the table into the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <table>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTBACK] <food_chicken> <freezer>", "[CLOSE] <freezer>", "[FIND] <food_carrot>", "[GRAB] <food_carrot>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_carrot> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'walllamp', 'bathroom', 'detergent', 'knifeblock', 'bathroom_counter', 'purse', 'alcohol', 'bookmark', 'stovefan', 'toaster', 'stereo', 'food_jam', 'dining_room', 'powersocket', 'food_food', 'mouthwash', 'tray', 'after_shave', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office', 'laser_pointer']\nTask:Walk through\nresponse:\n{\n    \"thoughts\": Walk through the hallway.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dining_room>', '[WALK] <home_office>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'crayon', 'freezer', 'bathroom', 'toothbrush', 'drawing', 'knifeblock', 'floor', 'phone', 'mail', 'brush', 'ceilinglamp', 'picture', 'window', 'couch', 'bowl', 'wall', 'pasta', 'bench', 'tape', 'newspaper', 'light_bulb', 'television', 'toothbrush_holder', 'shaving_cream', 'vacuum_cleaner', 'spoon', 'bathroom_cabinet', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I get home from work and walk into the living room. I see the couch and go to sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'tvstand', 'faucet', 'mousepad', 'closetdrawer', 'couch', 'food_jam', 'wall', 'nightstand', 'oven', 'newspaper', 'food_food', 'vacuum_cleaner', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'table', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I go to the bathroom, walk to the sink, turn on the faucet and then wash my hands under the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['toilet', 'dog', 'walllamp', 'nail_polish', 'ceilingfan', 'orchid', 'knife', 'character', 'mat', 'stereo', 'form', 'wall', 'cup', 'food_cheese', 'bed', 'food_food', 'video_game_controller', 'doorjamb', 'cards', 'hair', 'controller', 'conditioner', 'maindoor', 'laundry_detergent', 'door', 'napkin', 'light', 'vase', 'clothes_shirt', 'scrabble']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the TV remote. I change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'tea', 'band-aids', 'walllamp', 'nail_polish', 'knifeblock', 'bills', 'cpuscreen', 'hanger', 'shampoo', 'phone', 'floor', 'bedroom', 'desk', 'toaster', 'stereo', 'clothes_socks', 'clothes_pants', 'clothes_hat', 'bowl', 'dining_room', 'bench', 'newspaper', 'clothes_gloves', 'coffe_maker', 'curtain', 'diary', 'hairbrush', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I pull up my online classes on the computer to look at what the assigned reading is. I get my textbook, notebook, and mechanical pencil. As I'm reading, I take notes over important points. I then use those notes to complete the assignment for the week. Frequently that involves writing a post on an online discussion board that answers a question related to things from the reading.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['glue', 'dry_pasta', 'filing_cabinet', 'microwave', 'toilet', 'freezer', 'ground_coffee', 'walllamp', 'orchid', 'drawing', 'bills', 'bathroom_counter', 'floor', 'clothes_jacket', 'bookmark', 'ceilinglamp', 'mat', 'tvstand', 'floor_lamp', 'window', 'couch', 'wall', 'bed', 'novel', 'television', 'curtain', 'doorjamb', 'sink', 'table', 'board_game']\nTask:Read book\nresponse:\n", "output": {"thoughts": "First I walk into the living room, I find the lamp and switch it on. Then I find a book, grab it, sit in the sofa, open the book and read it. When I'm finished, I find a bookmark, grab it and put it in the book. Then I close the book, put it back, and switch off the lamp.", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>", "[FIND] <novel>", "[GRAB] <novel>", "[FIND] <couch>", "[SIT] <couch>", "[READ] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[PUTBACK] <bookmark> <novel>", "[PUTOBJBACK] <novel>", "[SWITCHOFF] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'freezer', 'hanger', 'floor', 'ceiling', 'cleaning_solution', 'mail', 'computer', 'toaster', 'mat', 'plate', 'faucet', 'brush', 'oil', 'candle', 'broom', 'wall', 'tape', 'bed', 'coffe_maker', 'ice', 'rag', 'console', 'kitchen_counter', 'cupboard', 'sink', 'bathroom_cabinet', 'colander']\nTask:Clean sink\nresponse:\n{\n    \"thoughts\": Squirt soap into the sink. Get a wet rag. Wipe out the sink. Rinse out the sink and the rag. Hang the rag out to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[FIND] <cleaning_solution>', '[GRAB] <cleaning_solution>', '[FIND] <cleaning_solution>', '[WALK] <sink>', '[POUR] <cleaning_solution> <sink>', '[PUTOBJBACK] <cleaning_solution>', '[WIPE] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <sink>', '[RINSE] <rag>', '[SWITCHOFF] <faucet>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'piano_bench', 'mat', 'shoe-shine_kit', 'balanceball', 'oil', 'clothes_hat', 'dining_room', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'mouthwash', 'doorjamb', 'towel_rack', 'vacuum_cleaner', 'door', 'home_office', 'light', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "switching on the light", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'blender', 'beer', 'toilet', 'detergent', 'food_butter', 'phone', 'bookshelf', 'remote_control', 'alcohol', 'button', 'table_cloth', 'piano_bench', 'toaster', 'tvstand', 'pillow', 'nightstand', 'pasta', 'wall_clock', 'tape', 'coffe_maker', 'doorjamb', 'cards', 'pencil', 'controller', 'clothes_dress', 'maindoor', 'sink', 'home_office', 'light']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I pick up the remote. I press button on remote.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <button>', '[PUSH] <button>']\n}\n", "input": "\nScene_Objects:['beer', 'microwave', 'trashcan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'keyboard', 'ceilinglamp', 'mat', 'picture', 'window', 'couch', 'dining_room', 'wall', 'oven', 'powersocket', 'toy', 'photoframe', 'television', 'oven_mitts', 'tray', 'centerpiece', 'kettle', 'towel_rack', 'pencil', 'dishrack', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to the living room and sit on the couch, ease into the sofa and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['nail_polish', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'faucet', 'couch', 'dining_room', 'wall', 'nightstand', 'bench', 'wall_clock', 'photoframe', 'coffe_maker', 'food_food', 'lighter', 'doorjamb', 'towel_rack', 'diary', 'shoes', 'comb', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go to the bathroom. turn on the water in the sink. apply soap to hands. rub hands together. rinse hands with water. turn off the water.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[RINSE] <hands_both>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'drawing', 'orchid', 'clothes_underwear', 'hanger', 'desk', 'clothes_jacket', 'mop_bucket', 'bookmark', 'shower', 'electric_shaver', 'clothes_socks', 'clothes_pants', 'food_kiwi', 'bench', 'basket_for_clothes', 'photoframe', 'bed', 'coffe_maker', 'food_rice', 'shaving_cream', 'dishrack', 'cupboard', 'clothes_dress', 'laundry_detergent', 'clothes_skirt', 'soap', 'board_game', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "I go to the bathroom, take the dirt clothes and put into the washing machine, then put the soap, set the cleaning program and switch on the machine", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <clothes_dress>", "[GRAB] <clothes_dress>", "[PUTBACK] <clothes_dress> <washing_machine>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[FIND] <clothes_skirt>", "[GRAB] <clothes_skirt>", "[PUTBACK] <clothes_skirt> <washing_machine>", "[FIND] <clothes_underwear>", "[GRAB] <clothes_underwear>", "[PUTBACK] <clothes_underwear> <washing_machine>", "[FIND] <clothes_socks>", "[GRAB] <clothes_socks>", "[PUTBACK] <clothes_socks> <washing_machine>", "[CLOSE] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'chair', 'freezer', 'band-aids', 'walllamp', 'nail_polish', 'orchid', 'drawing', 'cpuscreen', 'kitchen_cabinet', 'floor', 'keyboard', 'toaster', 'ceilinglamp', 'plate', 'stereo', 'mat', 'iron', 'food_peanut_butter', 'food_jam', 'wall', 'oven', 'check', 'food_food', 'food_rice', 'ice', 'food_onion', 'console', 'kitchen_counter', 'table']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the kitchen cabinet. Grab the plate and put it on the table.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <plate>', '[GRAB] <plate>', '[FIND] <table>', '[PUTBACK] <plate> <table>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'tablelamp', 'blender', 'microwave', 'toilet', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'shampoo', 'character', 'desk', 'keyboard', 'computer', 'mousepad', 'wall', 'bench', 'powersocket', 'photoframe', 'television', 'tray', 'mouse', 'dresser', 'controller', 'maindoor', 'home_office', 'clothes_skirt']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "Walk to bathroom, find toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'cup', 'newspaper', 'coffe_maker', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Start computer\nresponse:\n{\n    \"thoughts\": Sit in desk chair, press power button, let computer power on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'blender', 'walllamp', 'dough', 'knifeblock', 'hanger', 'floor', 'ceiling', 'character', 'keyboard', 'bookshelf', 'clothes_jacket', 'faucet', 'balanceball', 'pillow', 'wall', 'oven', 'food_food', 'wooden_spoon', 'curtain', 'kitchen_counter', 'headset', 'dishrack', 'cupboard', 'maindoor', 'sink', 'table', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Washing Hands", "actions": ["[WALK] <dining_room>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[TOUCH] <soap>", "[PUTOBJBACK] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_pizza', 'toilet', 'freezer', 'ground_coffee', 'bathroom', 'drawing', 'knifeblock', 'floor', 'ceiling', 'desk', 'bookshelf', 'alcohol', 'mat', 'shower', 'window', 'dining_room', 'food_kiwi', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'television', 'doorjamb', 'mouse', 'door', 'light', 'colander', 'coffee']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I get my groceries and go to the fridge. I open the fridge and put the groceries there.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'homework', 'mousepad', 'clothes_pants', 'wall', 'nightstand', 'photoframe', 'check', 'curtain', 'doorjamb', 'mouse', 'light', 'dirt']\nTask:Work\nresponse:\n", "output": {"thoughts": "Find a PC and do homework", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[FIND] <mouse>", "[FIND] <check>", "[FIND] <homework>", "[GRAB] <homework>", "[FIND] <homework>", "[READ] <homework>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'blender', 'food_salt', 'toilet', 'chair', 'toothbrush', 'drawing', 'orchid', 'bathroom_counter', 'bookshelf', 'purse', 'alcohol', 'clothes_jacket', 'ceilinglamp', 'foundation', 'pillow', 'bowl', 'dining_room', 'bed', 'curtain', 'lighter', 'doorjamb', 'towel_rack', 'console', 'chessboard', 'cd', 'clothes_scarf', 'instrument_guitar', 'light', 'teeth']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find a radio there and switch it on to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'floor', 'ceiling', 'desk', 'bedroom', 'stovefan', 'ceilinglamp', 'plate', 'mat', 'couch', 'wall', 'oven', 'food_food', 'tray', 'food_onion', 'curtain', 'diary', 'kitchen_counter', 'dishrack', 'sauce_pan', 'cupboard', 'spoon', 'home_office']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I pull a sauce pan from the cupboard. I turn the stove on. I take my ingredients from the pantry and fridge. I put the ingredients in the pan and I mix them with a spoon. I cook until the food is finished and transfer my food to a plate.", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[FIND] <pot>", "[GRAB] <pot>", "[FIND] <spoon>", "[GRAB] <spoon>", "[FIND] <sauce_pan>", "[PUTBACK] <spoon> <sauce_pan>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <plate>", "[PUTBACK] <pot> <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'microwave', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'folder', 'desk', 'alcohol', 'stovefan', 'plate', 'mousepad', 'wall', 'oven', 'food_food', 'check', 'tray', 'doorjamb', 'cards', 'mouse', 'kitchen_counter', 'shoes', 'cupboard', 'clothes_dress', 'sink', 'clothes_scarf', 'light', 'clothes_shirt', 'teeth']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk over and pick up the phone,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'chair', 'walllamp', 'trashcan', 'bills', 'knife', 'floor', 'desk', 'mail', 'button', 'toaster', 'plate', 'faucet', 'shower', 'foundation', 'couch', 'bowl', 'wall', 'nightstand', 'oven', 'bed', 'clothes_gloves', 'novel', 'coffe_maker', 'check', 'tray', 'sink', 'table', 'dishwasher', 'shoe_rack']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "In the kitchen, open up the dishwasher. Take plates from the sink and put them in the dishwasher one by one. Close the dishwasher. Program it. Push start.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'freezer', 'walllamp', 'cat', 'ceilingfan', 'drawing', 'cpuscreen', 'clothes_underwear', 'longboard', 'floor', 'ceiling', 'keyboard', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'oil', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'ice', 'curtain', 'after_shave', 'door', 'table', 'dishwasher', 'vase']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Sit on couch and watch the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['freezer', 'blow_dryer', 'ceilingfan', 'trashcan', 'toothbrush', 'drawing', 'knife', 'hanger', 'floor', 'ceiling', 'bookshelf', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'coffe_maker', 'television', 'food_donut', 'curtain', 'coffee_pot', 'controller', 'sink', 'bathroom_cabinet', 'table', 'soap', 'teeth', 'shoe_rack']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[TOUCH] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'trashcan', 'drawing', 'duster', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'mat', 'tvstand', 'picture', 'window', 'mousepad', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'powersocket', 'photoframe', 'food_food', 'curtain', 'doorjamb', 'sink', 'bathroom_cabinet', 'light', 'home_office', 'scrabble']\nTask:Dust\nresponse:\n{\n    \"thoughts\": Using the duster, lightly dust all hanging pictures on walls in living room. Dust all around surface of picture frame and front of pictures, moving duster back and forth, side to side, until all dust is gone.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <duster>', '[GRAB] <duster>', '[WALK] <wall>', '[WALK] <picture>', '[WIPE] <picture>', '[WALK] <picture>', '[WIPE] <picture>', '[WALK] <picture>', '[WIPE] <picture>', '[WALK] <filing_cabinet>', '[PUTBACK] <duster> <filing_cabinet>', '[CLOSE] <filing_cabinet>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'fork', 'beer', 'walllamp', 'bathroom', 'orchid', 'kitchen_cabinet', 'floor', 'phone', 'toaster', 'mat', 'plate', 'tvstand', 'window', 'dish_soap', 'couch', 'bowl', 'wall', 'oven', 'bench', 'coffe_maker', 'tray', 'rag', 'console', 'kitchen_counter', 'dishrack', 'hair', 'maindoor', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I went to the kitchen and collected the dishes and put them into the dish washer and wiped them with towel and put back into cabinet..", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>", "[GRAB] <plate>", "[WIPE] <plate>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[PUTOBJBACK] <dishrack>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'beer', 'chair', 'walllamp', 'bathroom', 'wallshelf', 'floor', 'ceiling', 'ceilinglamp', 'mat', 'stereo', 'couch', 'pillow', 'bowl', 'wall', 'nightstand', 'powersocket', 'cup', 'toilet_paper', 'television', 'after_shave', 'rag', 'doorjamb', 'coffee_pot', 'dishrack', 'instrument_guitar', 'table', 'soap', 'laser_pointer']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk to living room, walk to sofa, sit in sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['juice', 'detergent', 'knifeblock', 'cpuscreen', 'longboard', 'folder', 'character', 'bookshelf', 'table_cloth', 'toaster', 'stereo', 'picture', 'food_orange', 'nightstand', 'wall_clock', 'tape', 'light_bulb', 'coffe_maker', 'ice', 'tray', 'curtain', 'doorjamb', 'cd', 'shoes', 'dresser', 'door', 'sink', 'light', 'dishwasher', 'teeth']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find a radio there and switch it on to listen to music.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'mail', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'window', 'mousepad', 'wall', 'nightstand', 'photoframe', 'novel', 'doorjamb', 'mouse', 'light', 'dishwasher']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Switch on computer, Use keybord and mouse to open e-mail provider, type requested e-mail, click send, close e-mail provider, turn off computer,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[FIND] <keyboard>', '[FIND] <mail>', '[OPEN] <mail>', '[TYPE] <keyboard>', '[FIND] <button>', '[PUSH] <button>', '[CLOSE] <mail>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['water_glass', 'filing_cabinet', 'food_salt', 'microwave', 'freezer', 'walllamp', 'ground_coffee', 'knifeblock', 'needle', 'floor', 'mail', 'plate', 'faucet', 'window', 'pillow', 'nightstand', 'basket_for_clothes', 'cup', 'photoframe', 'bed', 'coffe_maker', 'coffee_filter', 'coin', 'water', 'after_shave', 'dresser', 'hairbrush', 'maindoor', 'table', 'shoe_rack']\nTask:Drink\nresponse:\n", "output": {"thoughts": "walk to kitchen, look at the water glass, grab that glass, find the water cane, open the water cane tap, drink water", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[TURNTO] <water_glass>", "[LOOKAT] <water_glass>", "[GRAB] <water_glass>", "[FIND] <water>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'microwave', 'ground_coffee', 'knifeblock', 'wallshelf', 'ceiling', 'phone', 'character', 'desk', 'piano_bench', 'stovefan', 'toaster', 'couch', 'clothes_hat', 'pillow', 'bench', 'oven', 'tape', 'cutting_board', 'doorjamb', 'cards', 'mouse', 'dishrack', 'cupboard', 'hairdryer', 'sink', 'clothes_scarf', 'table', 'dishwasher']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": i go find the phone. i pick up the phone,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <phone>', '[FIND] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'door', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I want to surf the internet and look around at funny pictures.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <keyboard>", "[GRAB] <keyboard>", "[FIND] <desk>", "[OPEN] <desk>", "[PUTBACK] <keyboard> <desk>", "[FIND] <mouse>", "[GRAB] <mouse>", "[FIND] <mousepad>", "[PUTBACK] <mouse> <mousepad>", "[PLUGIN] <keyboard>", "[PLUGIN] <mouse>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <chair>", "[SIT] <chair>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[GRAB] <mouse>", "[TOUCH] <mouse>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['bathroom', 'detergent', 'cat', 'knife', 'food_egg', 'character', 'stamp', 'bedroom', 'desk', 'mat', 'ceilinglamp', 'shower', 'faucet', 'stereo', 'window', 'oil', 'mousepad', 'clothes_hat', 'towel', 'pillow', 'dining_room', 'powersocket', 'coin', 'console', 'towel_rack', 'dishrack', 'cupboard', 'maindoor', 'sink', 'bag']\nTask:Keep cats out of room\nresponse:\n{\n    \"thoughts\": Stand in the doorway and create a loud noise so the cats will stay out.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'nail_polish', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mail', 'ceilinglamp', 'balanceball', 'mousepad', 'dustpan', 'towel', 'food_kiwi', 'wall', 'nightstand', 'novel', 'doorjamb', 'mouse', 'home_office', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "go to the office and use the computer to write and email", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mail>", "[OPEN] <mail>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'walllamp', 'bathroom', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'plate', 'window', 'toaster', 'clothes_pants', 'broom', 'powersocket', 'bench', 'bed', 'check', 'food_food', 'coffe_maker', 'after_shave', 'dresser', 'hair', 'door', 'sink', 'clothes_shirt']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": Get some clothes together that make an appropriate outfit.  Put them on,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <dresser>', '[FIND] <dresser>', '[OPEN] <dresser>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTON] <clothes_pants>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTON] <clothes_shirt>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tea', 'ground_coffee', 'walllamp', 'hanger', 'longboard', 'floor', 'ceiling', 'button', 'toaster', 'faucet', 'mousepad', 'candle', 'wall', 'nightstand', 'powersocket', 'keys', 'novel', 'coffe_maker', 'clothes_gloves', 'coffee_filter', 'water', 'ice', 'tray', 'food_onion', 'cupboard', 'door', 'light', 'dishwasher', 'soap']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'tablelamp', 'crayon', 'microwave', 'drawing', 'bills', 'cpuscreen', 'longboard', 'floor', 'bookshelf', 'computer', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'window', 'oil', 'couch', 'food_jam', 'food_orange', 'wall', 'bench', 'mouthwash', 'television', 'cupboard', 'table', 'clothes_skirt', 'light', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'ceilinglamp', 'laptop', 'faucet', 'tvstand', 'window', 'couch', 'microphone', 'towel', 'pillow', 'wall', 'oven', 'powersocket', 'photoframe', 'novel', 'food_food', 'television', 'tray', 'mouse', 'dresser', 'sink', 'clothes_scarf', 'home_office', 'table', 'clothes_skirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'food_salt', 'microwave', 'toilet', 'walllamp', 'drawing', 'needle', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'desk', 'computer', 'bookmark', 'stovefan', 'plate', 'tvstand', 'dish_soap', 'closetdrawer', 'dining_room', 'wall', 'wall_clock', 'oven', 'powersocket', 'curtain', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'alcohol', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'food_cheese', 'toothbrush_holder', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'bathroom_cabinet', 'light', 'home_office']\nTask:Work\nresponse:\n", "output": {"thoughts": "Type on a computer in the home office.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'food_ice_cream', 'measuring_cup', 'knife', 'orchid', 'cpuscreen', 'bathroom_counter', 'ceiling', 'phone', 'alcohol', 'computer', 'table_cloth', 'mat', 'deck_of_cards', 'stereo', 'window', 'pillow', 'powersocket', 'cup', 'newspaper', 'check', 'ice', 'after_shave', 'lighter', 'diary', 'cd', 'cupboard', 'maindoor', 'colander', 'board_game']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Find radio, turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'freezer', 'bathroom', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bookshelf', 'piano_bench', 'ceilinglamp', 'tvstand', 'faucet', 'clothes_socks', 'pillow', 'wall', 'nightstand', 'bed', 'newspaper', 'food_onion', 'doorjamb', 'towel_rack', 'shoes', 'dresser', 'hair', 'hairdryer', 'maindoor', 'spoon', 'table']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "In the kitchen where I left the bags of groceries, I grab cheese from the grocery bag, I walk over to the refrigerator, open it up, neatly put cheese in the fridge, Close the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'crayon', 'freezer', 'bathroom', 'trashcan', 'dough', 'knifeblock', 'orchid', 'floor', 'ceiling', 'clothes_jacket', 'brush', 'ceilinglamp', 'mat', 'balanceball', 'clothes_socks', 'couch', 'microphone', 'form', 'wall', 'food_food', 'check', 'oven_mitts', 'wine', 'mouse', 'cupboard', 'food_steak', 'spoon', 'table', 'tooth_paste']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Go kitchen and make something for eat,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>']\n}\n", "input": "\nScene_Objects:['water_glass', 'microwave', 'needle', 'kitchen_cabinet', 'desk', 'alcohol', 'mat', 'ceilinglamp', 'balanceball', 'window', 'couch', 'clothes_pants', 'dining_room', 'nightstand', 'wall', 'wall_clock', 'oven', 'toilet_paper', 'mouthwash', 'water', 'ice', 'tray', 'rag', 'doorjamb', 'towel_rack', 'coffee_pot', 'pencil', 'door', 'dishwasher', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Get a glass of water to drink", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'walllamp', 'food_carrot', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'food_bacon', 'ceilinglamp', 'mat', 'tvstand', 'stereo', 'mousepad', 'couch', 'wall', 'powersocket', 'toy', 'photoframe', 'mouse', 'controller', 'door', 'table', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": turn on the computer and browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_pizza', 'food_salt', 'jelly', 'band-aids', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'broom', 'wall', 'powersocket', 'doorjamb', 'lighter', 'door', 'sink', 'light', 'table', 'laser_pointer', 'clothes_shirt', 'creditcard']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go into my room, find the light,  then turn on the lights.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'toilet', 'chair', 'walllamp', 'detergent', 'food_carrot', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'bookmark', 'mat', 'window', 'mousepad', 'broom', 'wall', 'wall_clock', 'nightstand', 'novel', 'coin', 'lighter', 'centerpiece', 'mouse', 'dresser']\nTask:Read book\nresponse:\n{\n    \"thoughts\": find a book to read. sit down. read the book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['water_glass', 'pot', 'food_salt', 'mop', 'freezer', 'walllamp', 'detergent', 'knifeblock', 'floor', 'ceiling', 'purse', 'button', 'piano_bench', 'food_bacon', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'form', 'wall', 'bench', 'food_food', 'television', 'water', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'maindoor', 'table']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I need to go to Kitchen.  Get a cup.  Fill cup.  Drink", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <freezer>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'chair', 'drawing', 'bills', 'knife', 'floor', 'ceilinglamp', 'stereo', 'window', 'couch', 'food_jam', 'form', 'towel', 'dining_room', 'wall', 'basket_for_clothes', 'photoframe', 'novel', 'food_food', 'check', 'television', 'oven_mitts', 'razor', 'tray', 'pencil', 'dresser', 'pajamas', 'table', 'soap']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit and read,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[CLOSE] <filing_cabinet>', '[WALK] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'board_game', 'walllamp', 'bathroom', 'ceilingfan', 'drawing', 'longboard', 'phone', 'ceiling', 'dvd_player', 'bedroom', 'tea_bag', 'stamp', 'keyboard', 'stovefan', 'plate', 'picture', 'form', 'pillow', 'oven', 'television', 'centerpiece', 'shoes', 'cd', 'bag', 'dishwasher', 'soap', 'tooth_paste', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Find the cat.  Pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'form', 'pillow', 'wall', 'pasta', 'powersocket', 'photoframe', 'television', 'food_donut', 'curtain', 'doorjamb', 'chessboard', 'dresser', 'laundry_detergent', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'ceilingfan', 'drawing', 'knifeblock', 'knife', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'stereo', 'food_noodles', 'couch', 'dustpan', 'form', 'dining_room', 'wall', 'powersocket', 'oven', 'check', 'television', 'curtain', 'kitchen_counter', 'cd', 'table', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Go to the sofa and watch TV", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'bathroom', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'window', 'mousepad', 'wall', 'wall_clock', 'powersocket', 'oven', 'check', 'television', 'oven_mitts', 'doorjamb', 'mouse', 'kitchen_counter', 'sink', 'napkin', 'table', 'shoe_rack']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I will walk to the computer at the desk in my home office, sit in the chair, turn on the computer, open a web browser, type in the address, and view the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'computer', 'clothes_jacket', 'table_cloth', 'mat', 'food_kiwi', 'wall', 'powersocket', 'clothes_gloves', 'food_food', 'coffe_maker', 'doorjamb', 'coffee_pot', 'headset', 'door', 'hairdryer', 'light', 'tooth_paste']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk towards to the bedroom, and I open the door of the bedroom, and I switch on the lights inside the bedroom, the switch is beside the door.", "actions": ["[WALK] <bedroom>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'stovefan', 'wall_clock', 'powersocket', 'bed', 'food_food', 'television', 'wooden_spoon', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'sink', 'hairdryer', 'home_office', 'light', 'clothes_skirt', 'bathtub', 'bathroom_cabinet']\nTask:Read book\nresponse:\n{\n    \"thoughts\": pick up your book and read it,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'drawing', 'orchid', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'table_cloth', 'mat', 'ceilinglamp', 'shower', 'plate', 'wall', 'nightstand', 'powersocket', 'food_cheese', 'bed', 'coffe_maker', 'tray', 'curtain', 'towel_rack', 'dresser', 'controller', 'clothes_scarf', 'home_office', 'bathroom_cabinet', 'board_game', 'scrabble', 'teeth']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "walk to bathroom , find shower , take a shower", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <shower>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'clothes_dress', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I want to surf the internet and look around at funny pictures.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <keyboard>', '[GRAB] <keyboard>', '[FIND] <desk>', '[OPEN] <desk>', '[PUTBACK] <keyboard> <desk>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[PLUGIN] <keyboard>', '[PLUGIN] <mouse>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <chair>', '[SIT] <chair>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[GRAB] <mouse>', '[TOUCH] <mouse>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['pot', 'measuring_cup', 'walllamp', 'bathroom', 'trashcan', 'ceilingfan', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'character', 'toaster', 'mat', 'dish_soap', 'couch', 'wall', 'nightstand', 'oven', 'powersocket', 'bench', 'photoframe', 'coffe_maker', 'tray', 'doorjamb', 'shaving_cream', 'cupboard', 'food_steak', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the dishwasher and put dish soap. I close the dishwasher and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'scissors', 'mop', 'microwave', 'toilet', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'desk', 'bookshelf', 'tvstand', 'shower', 'stereo', 'mousepad', 'wall', 'powersocket', 'photoframe', 'bed', 'food_food', 'television', 'curtain', 'doorjamb', 'coffee_pot', 'chessboard', 'laundry_detergent', 'door', 'board_game']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": walk to bathroom , run to toilet,\n    \"actions\": ['[WALK] <bathroom>', '[RUN] <toilet>']\n}\n", "input": "\nScene_Objects:['chair', 'band-aids', 'trashcan', 'dough', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'shampoo', 'dvd_player', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'shower', 'mousepad', 'food_orange', 'wall', 'oven', 'keys', 'mouse', 'shoes', 'hair', 'dishwasher', 'vase', 'coffee']\nTask:Read book\nresponse:\n", "output": {"thoughts": "find a book to read. sit down. read the book.", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <chair>", "[SIT] <chair>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['mop', 'toilet', 'freezer', 'food_carrot', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'piano_bench', 'stovefan', 'mat', 'faucet', 'wall', 'wall_clock', 'oven', 'cutting_board', 'toilet_paper', 'video_game_controller', 'television', 'tray', 'doorjamb', 'towel_rack', 'coffee_pot', 'kitchen_counter', 'dresser', 'conditioner', 'sink', 'bathroom_cabinet', 'light', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Walk to bathroom.  Find mouthwash or turn on water.  Put mouthwash or water in mouth.  Swish around for 30 seconds.  Spit mouthwash out.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <teeth>', '[RINSE] <teeth>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'scissors', 'walllamp', 'detergent', 'drawing', 'clothes_underwear', 'floor', 'phone', 'character', 'button', 'toaster', 'plate', 'tvstand', 'dish_soap', 'oil', 'candle', 'wall', 'oven', 'tape', 'box', 'coffe_maker', 'food_food', 'check', 'tray', 'centerpiece', 'controller', 'light', 'dishwasher', 'soap']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk into kitchen, find dishwasher, put plates in dishwasher, put soap in dishwasher, turn on dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'ground_coffee', 'detergent', 'trashcan', 'orchid', 'drawing', 'bedroom', 'tea_bag', 'bookshelf', 'envelope', 'alcohol', 'computer', 'ceilinglamp', 'mat', 'shower', 'deck_of_cards', 'food_noodles', 'microphone', 'bowl', 'wall', 'bench', 'powersocket', 'photoframe', 'television', 'dresser', 'napkin', 'table', 'clothes_shirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": First I need to find the cat in the living room. Then I walk over to the cat and pet it gently.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['toilet', 'chair', 'freezer', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'character', 'desk', 'bedroom', 'bookshelf', 'envelope', 'computer', 'ceilinglamp', 'stereo', 'wall', 'pasta', 'cup', 'tray', 'doorjamb', 'dresser', 'cupboard', 'hair', 'clothes_dress', 'door', 'clothes_scarf', 'maindoor', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to the Living Room. Switch on 1 light. Walk to dining room. Switch on 1 light. Walk to Entrance hall. Switch on 1 light.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <dining_room>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'freezer', 'trashcan', 'drawing', 'bathroom_counter', 'ceiling', 'phone', 'thread', 'computer', 'stovefan', 'faucet', 'shower', 'food_jam', 'dining_room', 'wall', 'powersocket', 'tape', 'coffe_maker', 'food_food', 'mirror', 'doorjamb', 'mouse', 'paper_towel', 'dresser', 'cupboard', 'sink', 'clothes_scarf', 'bathroom_cabinet', 'home_office', 'dirt']\nTask:Clean  mirror\nresponse:\n{\n    \"thoughts\": Spray Windex or glass and mirror cleaner onto mirror. Use paper towel to wipe off cleaner until mirror is dry and has no streaks.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[CLOSE] <bathroom_cabinet>', '[WALK] <mirror>', '[WIPE] <mirror>', '[FIND] <trashcan>', '[PUTBACK] <paper_towel> <trashcan>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'jelly', 'band-aids', 'chair', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'shampoo', 'ceiling', 'desk', 'keyboard', 'purse', 'brush', 'toaster', 'shoe-shine_kit', 'plate', 'faucet', 'mousepad', 'towel', 'dining_room', 'wall', 'photoframe', 'newspaper', 'coffe_maker', 'toothbrush_holder', 'mouse', 'door', 'home_office']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. I walk over to the toilet. I sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'phone', 'tea_bag', 'computer', 'button', 'bookmark', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'foundation', 'couch', 'clothes_hat', 'wall', 'oven', 'bench', 'television', 'curtain', 'vacuum_cleaner', 'dresser', 'maindoor', 'spoon', 'home_office', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": walk to living room, find the television, switch on the television, find the sofa, walk to sofa, sit in sofa, watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'chair', 'bathroom', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'mousepad', 'dustpan', 'wall', 'nightstand', 'photoframe', 'wine', 'doorjamb', 'mouse', 'dresser', 'light', 'colander']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Go find a computer, turn it on, click google, and browse the internet.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'freezer', 'drawing', 'orchid', 'floor', 'ceiling', 'desk', 'bedroom', 'bookshelf', 'computer', 'mat', 'shoe-shine_kit', 'oil', 'pillow', 'wall', 'pasta', 'powersocket', 'cup', 'food_food', 'video_game_controller', 'toothbrush_holder', 'television', 'doorjamb', 'kettle', 'door', 'light', 'clothes_skirt', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on light in an area,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'bathroom', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'remote_control', 'purse', 'button', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'couch', 'dining_room', 'wall', 'bench', 'food_food', 'coffe_maker', 'television', 'toilet_paper', 'wooden_spoon', 'curtain', 'pencil', 'shoes', 'table', 'light', 'vase']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go the living room, look for my tv remote control, take the control and then sit down on the couch and turn on the TV and push the buttons on the remote control to change the channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'plate', 'faucet', 'shower', 'couch', 'pillow', 'dining_room', 'wall', 'pasta', 'newspaper', 'novel', 'check', 'toilet_paper', 'food_food', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'controller', 'door', 'table', 'light', 'dishwasher']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room. I cross the room to the couch. I sit down and relax.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'walllamp', 'knifeblock', 'orchid', 'cpuscreen', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'food_bacon', 'table_cloth', 'stovefan', 'faucet', 'dining_room', 'wall', 'oven', 'photoframe', 'novel', 'coffe_maker', 'drinking_glass', 'wine', 'shaving_cream', 'kitchen_counter', 'cupboard', 'maindoor', 'sink', 'bathroom_cabinet', 'dishwasher', 'scrabble']\nTask:Drink\nresponse:\n", "output": {"thoughts": "First I go into the kitchen, open the cupboard and find a glass.Then I close the cupboard. Then I go to the sink, turn the faucet on fill the glass under the faucet. I turn off the faucet and then I drink.", "actions": ["[WALK] <dining_room>", "[WALK] <cupboard>", "[FIND] <cupboard>", "[OPEN] <cupboard>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[PUTBACK] <drinking_glass> <sink>", "[GRAB] <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'stovefan', 'balanceball', 'mat', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'light', 'dishwasher']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to office, sit at desk, turn on computer, enter password, open application and begin work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[PUSH] <mouse>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['spectacles', 'food_bread', 'filing_cabinet', 'tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'mousepad', 'food_noodles', 'pillow', 'wall', 'nightstand', 'oven', 'basket_for_clothes', 'doorjamb', 'towel_rack', 'mouse', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "walk to home office ,find computer ,switch on computer and work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'microwave', 'trashcan', 'drawing', 'floor', 'dvd_player', 'desk', 'envelope', 'ceilinglamp', 'tvstand', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'cup', 'check', 'television', 'coin', 'colander', 'after_shave', 'cd', 'cupboard', 'table', 'vase', 'soap']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'cat', 'food_butter', 'orchid', 'hanger', 'floor', 'shampoo', 'phone', 'mat', 'shower', 'electric_shaver', 'dustpan', 'microphone', 'towel', 'pillow', 'dining_room', 'wall', 'bench', 'tape', 'food_cheese', 'check', 'wine', 'curtain', 'vacuum_cleaner', 'dresser', 'sink', 'napkin', 'bathroom_cabinet', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <home_office>", "[WALK] <address_book>", "[FIND] <address_book>", "[GRAB] <address_book>", "[READ] <address_book>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'computer', 'toaster', 'ceilinglamp', 'window', 'food_noodles', 'couch', 'microphone', 'pillow', 'wall', 'newspaper', 'novel', 'coffe_maker', 'television', 'coin', 'toothbrush_holder', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'bathroom_cabinet', 'table', 'vase', 'laser_pointer', 'creditcard']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Sit on couch. Grab book. Open book. Read book.,\n    \"actions\": ['[WALK] <couch>', '[SIT] <couch>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'chair', 'freezer', 'ground_coffee', 'ceilingfan', 'drawing', 'needle', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'food_bacon', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'clothes_pants', 'wall', 'toy', 'food_food', 'television', 'oven_mitts', 'curtain', 'diary', 'headset', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "I go in the living room. Then I find the remote, sit on the couch and turn on the television. Then I use the remote the change the TV channel.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'pot', 'filing_cabinet', 'microwave', 'chair', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'face', 'toaster', 'stovefan', 'faucet', 'oil', 'towel', 'wall', 'facial_cleanser', 'check', 'coffe_maker', 'tray', 'rag', 'mouse', 'kitchen_counter', 'dresser', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap']\nTask:Wash face\nresponse:\n{\n    \"thoughts\": Turn the faucet on. Wet the face. Take facial cleanser out of packaging. Lather on face. Rinse face with water. Turn off the faucet. Dry face with towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <face>', '[WASH] <face>', '[FIND] <rag>', '[GRAB] <rag>', '[RINSE] <rag>', '[FIND] <facial_cleanser>', '[GRAB] <facial_cleanser>', '[POUR] <facial_cleanser> <rag>', '[PUTOBJBACK] <facial_cleanser>', '[SCRUB] <face>', '[RINSE] <rag>', '[PUTOBJBACK] <rag>', '[RINSE] <face>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <face>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'detergent', 'food_butter', 'trashcan', 'hanger', 'floor', 'phone', 'ceiling', 'bedroom', 'computer', 'button', 'mat', 'laptop', 'plate', 'window', 'wall', 'nightstand', 'bed', 'check', 'curtain', 'doorjamb', 'dresser', 'cupboard', 'controller', 'door', 'light', 'dishwasher', 'laser_pointer', 'shoe_rack']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "first i will go to home office and switch on the lights and answer the call after then i will put back it and switch off the lights and close the door", "actions": ["[RUN] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[FIND] <phone>", "[GRAB] <phone>", "[PUTOBJBACK] <phone>", "[SWITCHOFF] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'toilet', 'chair', 'freezer', 'band-aids', 'walllamp', 'cat', 'drawing', 'orchid', 'floor', 'ceiling', 'phone', 'bedroom', 'ceilinglamp', 'tvstand', 'clothes_socks', 'couch', 'pillow', 'dining_room', 'wall', 'bench', 'clothes_gloves', 'novel', 'tray', 'doorjamb', 'kitchen_counter', 'door', 'table', 'light', 'dishwasher']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['food_bread', 'scissors', 'mop', 'chair', 'ceilingfan', 'milk', 'toothbrush', 'knifeblock', 'phone', 'bookshelf', 'stovefan', 'toaster', 'tvstand', 'shower', 'couch', 'food_peanut_butter', 'dustpan', 'oven', 'photoframe', 'clothes_gloves', 'bed', 'video_game_controller', 'tray', 'towel_rack', 'clothes_scarf', 'table', 'light', 'soap', 'tooth_paste', 'scrabble']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "before going to bed or when wake up earily morning brush teeth", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'freezer', 'bathroom', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'food_egg', 'bathroom_counter', 'shampoo', 'phone', 'mat', 'ceilinglamp', 'shower', 'electric_shaver', 'clothes_socks', 'couch', 'microphone', 'dining_room', 'food_kiwi', 'nightstand', 'wall', 'check', 'food_food', 'after_shave', 'mouse', 'kitchen_counter', 'cupboard', 'home_office', 'laser_pointer']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go home and it's dark so I can not see anything,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'measuring_cup', 'toilet', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'button', 'ceilinglamp', 'faucet', 'oil', 'teeth', 'clothes_hat', 'towel', 'wall', 'photoframe', 'bed', 'doorjamb', 'diary', 'kitchen_counter', 'dishrack', 'coffee', 'hairbrush', 'maindoor', 'sink', 'bathroom_cabinet', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom. Turn on the sink. Rinse hands. Apply soap and scrub your hands together. Rinse hands again. Turn off faucet. Dry hands on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>", "[PUTOBJBACK] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'blender', 'chair', 'freezer', 'walllamp', 'nail_polish', 'detergent', 'drawing', 'knife', 'wallshelf', 'bathroom_counter', 'floor', 'desk', 'computer', 'mat', 'ceilinglamp', 'plate', 'faucet', 'mousepad', 'candle', 'microphone', 'bowl', 'wall', 'nightstand', 'bench', 'bed', 'keys', 'food_food', 'table']\nTask:Eat\nresponse:\n{\n    \"thoughts\": I get my prepared dinner set on the dining table, sit down on chair, pick up utensils, eat food,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[PUTBACK] <food_food> <table>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <fork>', '[GRAB] <fork>', '[FIND] <knife>', '[GRAB] <knife>', '[FIND] <plate>', '[EAT] <plate>']\n}\n", "input": "\nScene_Objects:['blender', 'dog', 'band-aids', 'drawing', 'bathroom_counter', 'floor', 'longboard', 'bedroom', 'food_cake', 'computer', 'table_cloth', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'window', 'couch', 'dustpan', 'clothes_pants', 'bowl', 'wall', 'television', 'oven_mitts', 'after_shave', 'doorjamb', 'towel_rack', 'cards', 'shoes', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room. I bend down. I sit on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'check', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'light']\nTask:Research\nresponse:\n{\n    \"thoughts\": I read different articles online to gain a better understanding of a particular business, industry, method, terminology, etc.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <check>', '[GRAB] <check>', '[READ] <check>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'freezer', 'nail_polish', 'trashcan', 'bills', 'drawing', 'hanger', 'bathroom_counter', 'character', 'bedroom', 'keyboard', 'computer', 'stereo', 'shower', 'form', 'dining_room', 'powersocket', 'oven', 'bench', 'cup', 'coffe_maker', 'mouthwash', 'television', 'coin', 'wooden_spoon', 'tray', 'doorjamb', 'mouse', 'bathroom_cabinet']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I reach out arm. I put hand on radio. I turn on radio. I pick a station.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[TOUCH] <stereo>", "[TURNTO] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'freezer', 'bathroom', 'drawing', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bedroom', 'desk', 'bookshelf', 'shower', 'homework', 'wall', 'powersocket', 'box', 'bed', 'coffe_maker', 'television', 'after_shave', 'doorjamb', 'towel_rack', 'mouse', 'cards', 'cupboard', 'home_office', 'light', 'dishwasher']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk into living room. Find wall. Find light button. Switch on light button.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'food_ice_cream', 'toilet', 'chair', 'freezer', 'ceilingfan', 'drawing', 'longboard', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'mat', 'window', 'oil', 'couch', 'wall', 'bench', 'mouthwash', 'television', 'curtain', 'shaving_cream', 'mouse', 'headset', 'bathroom_cabinet', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "go to the living room. find the remote. sit on the couch. turn on the tv.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'juice', 'microwave', 'freezer', 'bills', 'floor', 'ceiling', 'desk', 'tea_bag', 'computer', 'ceilinglamp', 'plate', 'window', 'mousepad', 'couch', 'dining_room', 'broom', 'wall', 'powersocket', 'novel', 'food_food', 'mouthwash', 'television', 'curtain', 'rag', 'towel_rack', 'kitchen_counter', 'comb', 'door', 'laser_pointer']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'filing_cabinet', 'tablelamp', 'chair', 'freezer', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'purse', 'computer', 'mail', 'toaster', 'tvstand', 'mousepad', 'wall', 'check', 'ice', 'tray', 'doorjamb', 'towel_rack', 'coffee_pot', 'cards', 'dresser', 'cupboard', 'door', 'napkin', 'light', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk over to the phone. .I reach my arm out. I open my hand. I place my hand on the phone. I close my fingers around the phone. I pick up the phone.", "actions": ["[WALK] <phone>", "[TOUCH] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'ground_coffee', 'wallshelf', 'hanger', 'floor', 'stamp', 'desk', 'bookshelf', 'tea_bag', 'envelope', 'computer', 'mop_bucket', 'mat', 'shower', 'foundation', 'form', 'wall', 'powersocket', 'cutting_board', 'photoframe', 'novel', 'check', 'television', 'wooden_spoon', 'after_shave', 'doorjamb', 'door', 'light']\nTask:Read to child\nresponse:\n{\n    \"thoughts\": Read a book out loud to a child,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>', '[WALK] <filing_cabinet>', '[OPEN] <filing_cabinet>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[TURNTO] <novel>', '[LOOKAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[TURNTO] <novel>', '[POINTAT] <novel>', '[READ] <novel>', '[READ] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'freezer', 'food_butter', 'toothbrush', 'knifeblock', 'food_egg', 'cpuscreen', 'longboard', 'bathroom_counter', 'phone', 'ceiling', 'tea_bag', 'plate', 'shower', 'deck_of_cards', 'window', 'electric_shaver', 'dish_soap', 'broom', 'bench', 'powersocket', 'tape', 'mouthwash', 'centerpiece', 'shaving_cream', 'mouse', 'bathroom_cabinet', 'dishwasher', 'coffee']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Pick up dish.  Put in soapy water.  Wash dish.", "actions": ["[WALK] <dining_room>", "[WALK] <dish_soap>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <dish_soap> <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[WIPE] <plate>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'blender', 'beer', 'microwave', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'table_cloth', 'toaster', 'plate', 'faucet', 'tvstand', 'window', 'dish_soap', 'food_orange', 'wall', 'cup', 'bed', 'food_food', 'tray', 'wooden_spoon', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'sink']\nTask:Wash dirty dishes\nresponse:\n{\n    \"thoughts\": Wet the sponge, apply soap to sponge, scrub dishes, rinse dishes, place on dish rack.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <sponge> <sink>', '[GRAB] <sponge>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['tea', 'freezer', 'walllamp', 'knifeblock', 'kitchen_cabinet', 'floor', 'ceiling', 'dvd_player', 'desk', 'table_cloth', 'ceilinglamp', 'plate', 'mousepad', 'couch', 'candle', 'tape', 'light_bulb', 'novel', 'coin', 'water', 'drinking_glass', 'wine', 'tray', 'coffee_pot', 'mouse', 'cd', 'conditioner', 'maindoor', 'napkin', 'home_office']\nTask:Drink\nresponse:\n", "output": {"thoughts": "You walk to the kitchen and open the kitchen cabinet and grab a drinking glass. You find the water and pour the water into the drinking glass. You drink the water from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'dog', 'freezer', 'dough', 'knifeblock', 'knife', 'hanger', 'bathroom_counter', 'ceiling', 'toaster', 'tvstand', 'faucet', 'couch', 'pasta', 'powersocket', 'oven', 'photoframe', 'novel', 'television', 'wooden_spoon', 'curtain', 'coffee_pot', 'mouse', 'dresser', 'door', 'sink', 'home_office', 'light', 'bathroom_cabinet']\nTask:Greet guests\nresponse:\n{\n    \"thoughts\": Say hi,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <child>', '[FIND] <child>', '[GREET] <child>']\n}\n", "input": "\nScene_Objects:['sheets', 'scissors', 'mop', 'toilet', 'blow_dryer', 'detergent', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'toaster', 'tvstand', 'shower', 'mousepad', 'pillow', 'wall', 'powersocket', 'oven', 'cup', 'photoframe', 'bed', 'food_onion', 'doorjamb', 'shaving_cream', 'mouse', 'kitchen_counter', 'bathroom_cabinet']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to bathroom, I open the door of the bathroom, and sit on the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[SIT] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'tea', 'mop', 'microwave', 'toilet', 'toothbrush', 'drawing', 'knifeblock', 'orchid', 'phone', 'ceiling', 'computer', 'stovefan', 'toaster', 'tvstand', 'window', 'floor_lamp', 'couch', 'clothes_hat', 'photoframe', 'after_shave', 'food_onion', 'headset', 'chessboard', 'cupboard', 'controller', 'door', 'bathroom_cabinet', 'clothes_skirt']\nTask:Switch on lamp\nresponse:\n{\n    \"thoughts\": Press the button so the light comes on.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'food_carrot', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'mail', 'ceilinglamp', 'mat', 'iron', 'balanceball', 'faucet', 'food_orange', 'pillow', 'wall', 'pasta', 'box', 'cup', 'newspaper', 'food_food', 'check', 'toilet_paper', 'chessboard', 'maindoor', 'napkin', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Putting groceries in refridgerator", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'toilet', 'freezer', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'envelope', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'shower', 'stereo', 'mousepad', 'towel', 'wall', 'powersocket', 'photoframe', 'bed', 'toilet_paper', 'tray', 'curtain', 'doorjamb', 'dresser', 'controller', 'door', 'food_steak', 'light', 'tooth_paste']\nTask:Take shower\nresponse:\n{\n    \"thoughts\": walk to bathroom , find shower , take a shower,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <shower>', '[FIND] <shower>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_salt', 'chair', 'blow_dryer', 'bathroom', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'shampoo', 'bookshelf', 'purse', 'keyboard', 'ceilinglamp', 'deck_of_cards', 'shower', 'window', 'couch', 'wall', 'powersocket', 'tape', 'photoframe', 'television', 'food_rice', 'after_shave', 'controller', 'conditioner', 'door', 'table', 'scrabble']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and I lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_pizza', 'tablelamp', 'food_ice_cream', 'crayon', 'shampoo', 'keyboard', 'mail', 'bookmark', 'stovefan', 'faucet', 'pillow', 'dining_room', 'broom', 'wall', 'love_seat', 'photoframe', 'novel', 'food_food', 'wine', 'tray', 'after_shave', 'shaving_cream', 'kettle', 'headset', 'pajamas', 'laundry_detergent', 'door', 'bathroom_cabinet', 'shoe_rack']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Obtain a book from the book shelf. Sit in the easy chair.  Read the book.  Mark the stopping point with a bookmark.  Put the book back on the book shelf.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <love_seat>', '[SIT] <love_seat>', '[READ] <novel>', '[FIND] <bookmark>', '[GRAB] <bookmark>', '[PUTBACK] <bookmark> <novel>', '[PUTOBJBACK] <novel>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'beer', 'jelly', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'keyboard', 'clothes_jacket', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'food_noodles', 'couch', 'wall', 'oven', 'basket_for_clothes', 'cup', 'television', 'curtain', 'conditioner', 'maindoor', 'table', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room. Sit on the couch. Turn on TV.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_ice_cream', 'measuring_cup', 'toilet', 'chair', 'drawing', 'food_egg', 'clothes_underwear', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'stamp', 'bookshelf', 'button', 'shoe-shine_kit', 'shower', 'window', 'broom', 'wall', 'wall_clock', 'powersocket', 'bench', 'standingmirror', 'novel', 'coin', 'doorjamb', 'door', 'light', 'table', 'creditcard']\nTask:Read\nresponse:\n{\n    \"thoughts\": We pick out a book from the bookshelf and read it.,\n    \"actions\": ['[WALK] <novel>', '[WALK] <bookshelf>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'food_salt', 'freezer', 'walllamp', 'trashcan', 'food_carrot', 'knifeblock', 'floor', 'ceiling', 'food_bacon', 'stovefan', 'ceilinglamp', 'toaster', 'mousepad', 'wall', 'nightstand', 'oven', 'toy', 'box', 'photoframe', 'keys', 'toothbrush_holder', 'tray', 'kitchen_counter', 'hair', 'door', 'sink', 'dishwasher']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "go to the kitchen and cook some thing to eat", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <oven>", "[OPEN] <oven>", "[FIND] <food_pizza>", "[GRAB] <food_pizza>", "[PUTIN] <food_pizza> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['band-aids', 'trashcan', 'wallshelf', 'clothes_underwear', 'desk', 'keyboard', 'mail', 'computer', 'tvstand', 'shower', 'window', 'clothes_socks', 'couch', 'food_peanut_butter', 'pillow', 'basket_for_clothes', 'keys', 'coffe_maker', 'television', 'food_donut', 'tray', 'rag', 'headset', 'clothes_scarf', 'bathtub', 'sink', 'light', 'dishwasher', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'freezer', 'bathroom', 'cpuscreen', 'desk', 'tea_bag', 'alcohol', 'tvstand', 'oven', 'tape', 'cutting_board', 'cup', 'clothes_gloves', 'novel', 'toilet_paper', 'television', 'coffee_filter', 'wooden_spoon', 'mouse', 'cd', 'controller', 'maindoor', 'sink', 'home_office', 'light', 'dishwasher', 'soap', 'dirt', 'shoe_rack']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk into living room. Turn on radio.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'toilet', 'chair', 'freezer', 'food_butter', 'food_carrot', 'drawing', 'floor', 'ceiling', 'keyboard', 'mat', 'ceilinglamp', 'shower', 'oil', 'food_peanut_butter', 'food_orange', 'nightstand', 'wall', 'powersocket', 'photoframe', 'food_food', 'food_onion', 'doorjamb', 'pencil', 'dresser', 'door', 'napkin', 'light', 'shoe_rack']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": put away groceries in fridge. open the door, put on shelf, close the door.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'couch', 'wall', 'nightstand', 'toy', 'photoframe', 'curtain', 'rag', 'doorjamb', 'mouse', 'dresser', 'door', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "walk into Living room, look for computer, walk to computer, turn on computer, grab mouse, click mouse", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[WALK] <computer>", "[TURNTO] <computer>", "[FIND] <mouse>", "[GRAB] <mouse>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'chair', 'detergent', 'food_carrot', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'bookmark', 'toaster', 'balanceball', 'wall', 'powersocket', 'photoframe', 'coffee_filter', 'tray', 'doorjamb', 'cards', 'mouse', 'shaving_cream', 'shoes', 'dresser', 'cupboard', 'door', 'napkin', 'home_office', 'light', 'vase']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": switching on light in an area,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['glue', 'scissors', 'tablelamp', 'freezer', 'trashcan', 'drawing', 'floor', 'shampoo', 'folder', 'bookshelf', 'purse', 'computer', 'window', 'wall', 'powersocket', 'cutting_board', 'cup', 'bed', 'photoframe', 'television', 'curtain', 'food_onion', 'shaving_cream', 'mouse', 'pencil', 'hair', 'maindoor', 'clothes_scarf', 'home_office', 'colander']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "go t the bedroom and go to sleep", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[FIND] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'walllamp', 'cat', 'drawing', 'orchid', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'character', 'envelope', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'wall_clock', 'oven', 'coffe_maker', 'television', 'curtain', 'hair', 'bathroom_cabinet', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I would walk into the living room. Then sit on the couch and pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'toilet', 'jelly', 'ceilingfan', 'drawing', 'hanger', 'floor', 'stamp', 'desk', 'ceilinglamp', 'balanceball', 'faucet', 'window', 'couch', 'clothes_pants', 'pillow', 'wall', 'nightstand', 'bed', 'photoframe', 'television', 'oven_mitts', 'food_rice', 'doorjamb', 'cards', 'shaving_cream', 'sink', 'table', 'light', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "go to living room and lie down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'orchid', 'floor', 'bedroom', 'bookshelf', 'toaster', 'tvstand', 'faucet', 'shower', 'stereo', 'closetdrawer', 'couch', 'broom', 'wall', 'wall_clock', 'photoframe', 'bed', 'television', 'mouse', 'kitchen_counter', 'comb', 'dresser', 'sink', 'spoon', 'clothes_skirt']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk into the bath room and go to the toilet,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'ground_coffee', 'food_butter', 'toothbrush', 'knifeblock', 'drawing', 'phone', 'keyboard', 'bookshelf', 'table_cloth', 'mat', 'plate', 'electric_shaver', 'window', 'pillow', 'nightstand', 'oven', 'newspaper', 'clothes_gloves', 'photoframe', 'toilet_paper', 'water', 'centerpiece', 'towel_rack', 'kettle', 'pencil', 'clothes_scarf', 'tooth_paste', 'teeth']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "Walk to bathroom. Find tooth paste, tooth brush and water. Squeeze tooth paste on tooth brush. Wash teeth.", "actions": ["[WALK] <bathroom>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[FIND] <toothbrush>", "[FIND] <water>", "[SQUEEZE] <tooth_paste>", "[FIND] <teeth>", "[WASH] <teeth>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'blender', 'beer', 'microwave', 'knifeblock', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'table_cloth', 'toaster', 'plate', 'faucet', 'tvstand', 'window', 'dish_soap', 'food_orange', 'wall', 'cup', 'bed', 'food_food', 'tray', 'wooden_spoon', 'mouse', 'pencil', 'kitchen_counter', 'dishrack', 'cupboard', 'sink']\nTask:Wash dirty dishes\nresponse:\n{\n    \"thoughts\": Wet the sponge, apply soap to sponge, scrub dishes, rinse dishes, place on dish rack.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sponge>', '[GRAB] <sponge>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[PUTBACK] <sponge> <sink>', '[GRAB] <sponge>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[FIND] <dishrack>', '[PUTBACK] <plate> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[RINSE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['food_bread', 'tablelamp', 'walllamp', 'ground_coffee', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'mat', 'faucet', 'couch', 'pillow', 'wall', 'pasta', 'wall_clock', 'cup', 'photoframe', 'bed', 'television', 'water', 'wine', 'curtain', 'paper_towel', 'sink', 'bathroom_cabinet', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk into the bathroom. I turn on the water. I put soap on my hands. I rinse my hands. I turn off the water. I dry my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <water>", "[TURNTO] <water>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[TURNTO] <water>", "[FIND] <paper_towel>", "[GRAB] <paper_towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['scissors', 'tablelamp', 'chair', 'freezer', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'photoframe', 'newspaper', 'television', 'doorjamb', 'dresser', 'soap', 'instrument_violin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit on my couch and watch TV in my living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[TOUCH] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[LOOKAT] <television>', '[WATCH] <television>', '[WATCH] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['spectacles', 'pot', 'tablelamp', 'chair', 'dog', 'walllamp', 'ceilingfan', 'knifeblock', 'needle', 'bathroom_counter', 'floor', 'computer', 'shoe-shine_kit', 'dustpan', 'box', 'cup', 'bed', 'keys', 'television', 'food_donut', 'tray', 'curtain', 'doorjamb', 'kettle', 'kitchen_counter', 'shoes', 'comb', 'sink', 'bathroom_cabinet', 'board_game']\nTask:Brush teeth\nresponse:\n", "output": {"thoughts": "go to the bathroom and brush your teeth", "actions": ["[WALK] <bathroom>", "[WALK] <toothbrush>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[WALK] <tooth_paste>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[POUR] <tooth_paste> <toothbrush>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'knifeblock', 'cpuscreen', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'ceilinglamp', 'faucet', 'couch', 'dustpan', 'pillow', 'towel', 'wall', 'nightstand', 'oven', 'box', 'bed', 'keys', 'check', 'toilet_paper', 'television', 'kitchen_counter', 'dresser', 'instrument_guitar', 'soap', 'tooth_paste']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": Go to living room.  Find sofa.  Sit on sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'jelly', 'ceilingfan', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'floor', 'bookshelf', 'computer', 'clothes_jacket', 'piano_bench', 'stovefan', 'toaster', 'tvstand', 'ceilinglamp', 'balanceball', 'stereo', 'brush', 'food_jam', 'clothes_hat', 'nightstand', 'pasta', 'powersocket', 'bench', 'bed', 'oven_mitts', 'tray', 'towel_rack', 'home_office']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to the radio.  I switch on the radio and listen", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'bookshelf', 'remote_control', 'envelope', 'clothes_jacket', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'bed', 'television', 'curtain', 'doorjamb', 'door', 'light', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Enter the living room, find TV remote, grab TV remote, switch on TV, find sofa, sit down on the sofa, watch TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['sponge', 'chair', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'remote_control', 'food_cake', 'envelope', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'couch', 'wall', 'wall_clock', 'light_bulb', 'television', 'curtain', 'hair', 'food_steak', 'sink', 'bathroom_cabinet', 'table', 'dishwasher']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I enter the living room and sit on the couch. I grab the remote and turn on the TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'envelope', 'computer', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'photoframe', 'check', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I go to the home office to work on my project. I turn on the light and walk to my desk. I turn on the computer. Then I look through the documents. After I'm done, I turn off the computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[WALK] <desk>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <check>', '[TURNTO] <check>', '[LOOKAT] <check>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['freezer', 'trashcan', 'food_carrot', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'alcohol', 'button', 'ceilinglamp', 'plate', 'clothes_socks', 'couch', 'food_orange', 'wall', 'bench', 'powersocket', 'photoframe', 'light_bulb', 'food_food', 'coffee_filter', 'drinking_glass', 'water', 'towel_rack', 'dishrack', 'cupboard', 'clothes_dress', 'light', 'colander']\nTask:Drink\nresponse:\n", "output": {"thoughts": "im feeling thirsty im going to kitchen and open the freezer where i can find some cold water and im going to drink it to end up my thirst", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <drinking_glass>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'filing_cabinet', 'food_bread', 'microwave', 'trashcan', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'bathroom_counter', 'bedroom', 'tea_bag', 'tvstand', 'window', 'form', 'pillow', 'broom', 'bench', 'basket_for_clothes', 'newspaper', 'bed', 'check', 'coffe_maker', 'tray', 'door', 'home_office', 'light', 'bathroom_cabinet', 'tooth_paste', 'coffee']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": I get dressed in my bedroom in the mornings.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'microwave', 'bathroom', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'toaster', 'mat', 'faucet', 'towel', 'wall', 'bench', 'photoframe', 'light_bulb', 'lighter', 'pencil', 'chessboard', 'dresser', 'hair', 'cupboard', 'sink', 'bathroom_cabinet', 'vase', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the sink, turn on the water,  put soap on my hands, rub them together, then rinse them off and turn off the water. I then towel off my hands.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[FIND] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'trashcan', 'cpuscreen', 'clothes_underwear', 'ceiling', 'folder', 'desk', 'computer', 'stovefan', 'shower', 'mousepad', 'candle', 'clothes_hat', 'dining_room', 'newspaper', 'photoframe', 'food_food', 'video_game_controller', 'oven_mitts', 'curtain', 'pencil', 'kitchen_counter', 'dishrack', 'laundry_detergent', 'door', 'light', 'dishwasher', 'soap', 'tooth_paste', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washer. Put clothes in washer. Put soap in washer. Close washer. Turn washer on.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_underwear>', '[GRAB] <clothes_underwear>', '[PUTBACK] <clothes_underwear> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'bathroom', 'milk', 'knifeblock', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'bookshelf', 'keyboard', 'purse', 'computer', 'shower', 'window', 'electric_shaver', 'mousepad', 'oil', 'couch', 'form', 'wall', 'food_food', 'television', 'rag', 'lighter', 'mouse', 'controller', 'sink']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "first when i hear the ringing sound i will run to my living room and picks up and i will say hello", "actions": ["[RUN] <home_office>", "[WALK] <chair>", "[FIND] <chair>", "[SIT] <chair>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'food_ice_cream', 'toilet', 'freezer', 'drawing', 'floor', 'ceiling', 'phone', 'stovefan', 'mat', 'ceilinglamp', 'shower', 'bowl', 'nightstand', 'wall', 'powersocket', 'oven', 'photoframe', 'bed', 'food_food', 'coffe_maker', 'wooden_spoon', 'doorjamb', 'pencil', 'comb', 'door', 'light', 'soap', 'scrabble', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": In the kitchen where I left the bags of groceries, I grab cheese from the grocery bag, I walk over to the refrigerator, open it up, neatly put cheese in the fridge, Close the fridge,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <food_food>', '[FIND] <food_food>', '[GRAB] <food_food>', '[WALK] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['water_glass', 'chair', 'cat', 'hanger', 'bathroom_counter', 'floor', 'bookshelf', 'keyboard', 'ceilinglamp', 'shower', 'foundation', 'couch', 'food_orange', 'powersocket', 'bench', 'clothes_gloves', 'toilet_paper', 'razor', 'food_donut', 'mouse', 'kitchen_counter', 'shoes', 'dishrack', 'dresser', 'pajamas', 'laundry_detergent', 'clothes_scarf', 'home_office', 'laser_pointer', 'creditcard']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Pour water into a glass and drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'bathroom', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'mousepad', 'dustpan', 'wall', 'nightstand', 'photoframe', 'wine', 'doorjamb', 'mouse', 'dresser', 'light', 'colander']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Go find a computer, turn it on, click google, and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'chair', 'freezer', 'detergent', 'trashcan', 'dough', 'clothes_underwear', 'folder', 'stamp', 'bookshelf', 'clothes_jacket', 'table_cloth', 'mousepad', 'couch', 'food_jam', 'clothes_hat', 'pillow', 'nightstand', 'bench', 'oven', 'cutting_board', 'bed', 'food_food', 'check', 'doorjamb', 'console', 'cupboard', 'home_office', 'table']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "i go find the phone. i pick up the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['measuring_cup', 'chair', 'walllamp', 'trashcan', 'orchid', 'drawing', 'knifeblock', 'hanger', 'bathroom_counter', 'ceiling', 'bedroom', 'bookshelf', 'clothes_jacket', 'faucet', 'couch', 'clothes_pants', 'form', 'wall', 'cutting_board', 'photoframe', 'check', 'food_food', 'curtain', 'doorjamb', 'towel_rack', 'clothes_dress', 'door', 'sink', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open cabinet and collect clothes and put them into washing machine and wash clothes,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <clothes_dress>', '[GRAB] <clothes_dress>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <clothes_dress> <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_jacket>', '[GRAB] <clothes_jacket>', '[PUTBACK] <clothes_jacket> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['fork', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'milk', 'floor', 'ceiling', 'shampoo', 'purse', 'ceilinglamp', 'mat', 'faucet', 'oil', 'wall', 'pasta', 'tape', 'novel', 'food_food', 'food_donut', 'wooden_spoon', 'mouse', 'kitchen_counter', 'door', 'home_office', 'bathroom_cabinet', 'vase', 'soap', 'instrument_violin', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'food_pizza', 'pot', 'microwave', 'bathroom', 'bills', 'knifeblock', 'cpuscreen', 'drawing', 'longboard', 'floor', 'phone', 'folder', 'bedroom', 'keyboard', 'mail', 'food_bacon', 'piano_bench', 'button', 'table_cloth', 'shower', 'couch', 'dining_room', 'wall', 'powersocket', 'cutting_board', 'water', 'tray', 'pencil', 'door']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Drinking a glass of water,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'crayon', 'bathroom', 'drawing', 'floor', 'phone', 'character', 'keyboard', 'remote_control', 'button', 'ceilinglamp', 'shower', 'window', 'foundation', 'mousepad', 'couch', 'pillow', 'wall', 'oven', 'check', 'food_food', 'television', 'dresser', 'sink', 'table', 'dishwasher', 'clothes_skirt', 'tooth_paste', 'dirt']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "go to living room. grab remote control. sit on couch. point remote at tv. push buttons.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <couch>", "[SIT] <couch>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'microwave', 'needle', 'kitchen_cabinet', 'desk', 'alcohol', 'mat', 'ceilinglamp', 'balanceball', 'window', 'couch', 'clothes_pants', 'dining_room', 'nightstand', 'wall', 'wall_clock', 'oven', 'toilet_paper', 'mouthwash', 'water', 'ice', 'tray', 'rag', 'doorjamb', 'towel_rack', 'coffee_pot', 'pencil', 'door', 'dishwasher', 'vase']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Get a glass of water to drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <water_glass>', '[POUR] <water> <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'scissors', 'chair', 'freezer', 'dog', 'trashcan', 'ceilingfan', 'drawing', 'bills', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'faucet', 'mousepad', 'wall', 'nightstand', 'television', 'doorjamb', 'cards', 'mouse', 'light', 'table']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to my desk. I turn on the computer. I open a web page and look at it.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Surf web for money legitimate making opportunities\nresponse:\n{\n    \"thoughts\": Sit down in front of computer. Turn computer on. Go online. Surf the web.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'measuring_cup', 'freezer', 'ceilingfan', 'cpuscreen', 'longboard', 'ceiling', 'character', 'bedroom', 'toaster', 'stereo', 'balanceball', 'oil', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'nightstand', 'photoframe', 'coffe_maker', 'rag', 'doorjamb', 'towel_rack', 'shoes', 'controller', 'napkin', 'bathroom_cabinet', 'laser_pointer', 'dirt']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Go to the living room.  Find the telephone.  Lift up the receiver.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'toilet', 'chair', 'walllamp', 'detergent', 'trashcan', 'food_carrot', 'knifeblock', 'drawing', 'cpuscreen', 'knife', 'bathroom_counter', 'desk', 'keyboard', 'bookshelf', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'mousepad', 'pillow', 'dining_room', 'oven', 'photoframe', 'food_food', 'kitchen_counter', 'dresser', 'sink', 'home_office']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Walk to the lamp.  Find the switch.  Click the switch off.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[SWITCHOFF] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['mop', 'bathroom', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'remote_control', 'computer', 'clothes_jacket', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'food_noodles', 'couch', 'clothes_pants', 'pillow', 'wall', 'oven', 'check', 'television', 'tray', 'curtain', 'sink', 'table', 'light', 'teeth']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "You walk to the living room and sit on the sofa. You grab the remote control, look at it and point at the television, you watch television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[TURNTO] <remote_control>", "[LOOKAT] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'pot', 'food_salt', 'jelly', 'walllamp', 'detergent', 'knife', 'drawing', 'hanger', 'floor', 'dvd_player', 'character', 'desk', 'bedroom', 'plate', 'dish_soap', 'oil', 'wall', 'powersocket', 'bed', 'light_bulb', 'coffe_maker', 'toilet_paper', 'rag', 'doorjamb', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put dishes in the dishwasher and start the dishwasher.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[WALK] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'needle', 'cpuscreen', 'floor', 'character', 'keyboard', 'table_cloth', 'plate', 'faucet', 'window', 'oil', 'mousepad', 'pillow', 'wall', 'oven', 'food_cheese', 'bed', 'tray', 'kettle', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'light']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[TURNTO] <plate>", "[LOOKAT] <plate>", "[GRAB] <plate>", "[SCRUB] <plate>", "[RINSE] <plate>", "[WASH] <plate>", "[PUTBACK] <plate> <kitchen_counter>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'microwave', 'ceilingfan', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'food_cake', 'thread', 'faucet', 'towel', 'dining_room', 'wall', 'nightstand', 'bench', 'photoframe', 'water', 'rag', 'lighter', 'cards', 'console', 'mouse', 'conditioner', 'sink', 'bathroom_cabinet', 'table', 'soap', 'laser_pointer', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Go into the bathroom and go over to the sink. Grab soap and turn on the water. Wet your hands and rub them together so that the soap lathers. Rub the soap all over your hands thoroughly. Rinse off the soap with water and turn off the sink. Dry your hands with a towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <hands_both>', '[POUR] <water> <hands_both>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'purse', 'computer', 'ceilinglamp', 'mat', 'balanceball', 'electric_shaver', 'window', 'mousepad', 'dustpan', 'pillow', 'wall', 'nightstand', 'photoframe', 'food_food', 'doorjamb', 'mouse', 'door', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "turn on the computer and browse the internet", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'water_glass', 'pot', 'glue', 'walllamp', 'trashcan', 'knifeblock', 'ceiling', 'clothes_jacket', 'stovefan', 'toaster', 'mat', 'faucet', 'microphone', 'wall', 'cup', 'bed', 'food_food', 'check', 'water', 'food_donut', 'curtain', 'kitchen_counter', 'chessboard', 'cupboard', 'laundry_detergent', 'sink', 'spoon', 'bathroom_cabinet', 'dirt']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. open the cupboard. Find the water glass, grab it. Fill the glass with water and drink water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'tea', 'toilet', 'chair', 'detergent', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'mail', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'novel', 'food_onion', 'doorjamb', 'mouse', 'napkin', 'light', 'clothes_skirt']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "I go to the living room, where my computer is. I locate it, switch it on and type desired URL adress in the web browser.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'drawing', 'knife', 'wallshelf', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'box', 'photoframe', 'television', 'coffee_filter', 'after_shave', 'doorjamb', 'diary', 'kitchen_counter', 'dresser']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": Walk to livingroom, grab remote and change the channel on television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'chair', 'nail_polish', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'ceilinglamp', 'balanceball', 'shower', 'mousepad', 'pillow', 'wall', 'nightstand', 'basket_for_clothes', 'toothbrush_holder', 'doorjamb', 'mouse', 'light', 'vase']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to the Home Office.  Find the computer.  Turn on the computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'microwave', 'toothbrush', 'drawing', 'orchid', 'wallshelf', 'kitchen_cabinet', 'floor', 'ceiling', 'desk', 'bedroom', 'stovefan', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'mousepad', 'pillow', 'wall', 'pasta', 'oven', 'coffe_maker', 'water', 'tray', 'towel_rack', 'kitchen_counter', 'headset', 'sauce_pan', 'cupboard']\nTask:Cook some food\nresponse:\n{\n    \"thoughts\": Cook some pasta on the stove,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_cabinet>', '[FIND] <kitchen_cabinet>', '[OPEN] <kitchen_cabinet>', '[FIND] <pasta>', '[GRAB] <pasta>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <sauce_pan>', '[POUR] <water> <sauce_pan>', '[FIND] <oven>', '[SWITCHON] <oven>', '[PUTBACK] <pasta> <sauce_pan>', '[FIND] <sauce_pan>', '[GRAB] <sauce_pan>', '[PUTBACK] <sauce_pan> <oven>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'toaster', 'ceilinglamp', 'stovefan', 'balanceball', 'mat', 'mousepad', 'oil', 'wall', 'nightstand', 'photoframe', 'doorjamb', 'mouse', 'light', 'dishwasher']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to office, sit at desk, turn on computer, enter password, open application and begin work", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[PUSH] <mouse>", "[TURNTO] <computer>", "[LOOKAT] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'crayon', 'toilet', 'chair', 'bathroom', 'orchid', 'floor', 'ceiling', 'character', 'bedroom', 'desk', 'computer', 'tvstand', 'faucet', 'stereo', 'food_orange', 'photoframe', 'novel', 'television', 'curtain', 'doorjamb', 'diary', 'towel_rack', 'mouse', 'shoes', 'dresser', 'cupboard', 'sink', 'bathtub', 'home_office']\nTask:Get dressed\nresponse:\n{\n    \"thoughts\": I get dressed in my bedroom in the mornings.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PUTON] <shoes>']\n}\n", "input": "\nScene_Objects:['sponge', 'sheets', 'filing_cabinet', 'food_bread', 'toilet', 'ground_coffee', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'desk', 'keyboard', 'alcohol', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'plate', 'iron', 'mousepad', 'wall', 'oven', 'photoframe', 'light_bulb', 'coffe_maker', 'centerpiece', 'mouse', 'dresser', 'maindoor']\nTask:Go to toilet\nresponse:\n", "output": {"thoughts": "I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.", "actions": ["[WALK] <bathroom>", "[WALK] <toilet>", "[FIND] <toilet>", "[TURNTO] <toilet>", "[LOOKAT] <toilet>", "[WALK] <toilet>", "[TURNTO] <toilet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'detergent', 'trashcan', 'drawing', 'wallshelf', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'computer', 'toaster', 'mat', 'tvstand', 'shower', 'closetdrawer', 'wall', 'oven', 'cup', 'novel', 'water', 'drinking_glass', 'doorjamb', 'mouse', 'kitchen_counter', 'comb', 'cupboard', 'door']\nTask:Drink\nresponse:\n{\n    \"thoughts\": go to the kitchen,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[FIND] <water>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['pot', 'freezer', 'ceilingfan', 'dough', 'milk', 'knifeblock', 'bills', 'floor', 'ceiling', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'mop_bucket', 'ceilinglamp', 'microphone', 'wall', 'bench', 'powersocket', 'clothes_gloves', 'light_bulb', 'food_food', 'coffe_maker', 'tray', 'curtain', 'shaving_cream', 'controller', 'door', 'maindoor', 'dirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen and put groceries in fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[WALK] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'sheets', 'filing_cabinet', 'food_salt', 'toilet', 'band-aids', 'walllamp', 'wallshelf', 'cpuscreen', 'hanger', 'floor', 'phone', 'computer', 'mop_bucket', 'stovefan', 'tvstand', 'deck_of_cards', 'closetdrawer', 'newspaper', 'coffe_maker', 'doorjamb', 'dresser', 'cupboard', 'door', 'food_steak', 'bag', 'sink', 'table', 'dishwasher', 'bathroom_cabinet']\nTask:Put umbrella away\nresponse:\n{\n    \"thoughts\": Place the umbrella in the umbrella bucket point first.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mop_bucket>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'freezer', 'blow_dryer', 'trashcan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'bookmark', 'ceilinglamp', 'mat', 'window', 'couch', 'microphone', 'clothes_hat', 'pillow', 'wall', 'bench', 'cup', 'keys', 'television', 'towel_rack', 'cd', 'cupboard', 'conditioner', 'door', 'home_office', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to the living room and relax on the sofa with a book", "actions": ["[RUN] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[WALK] <couch>", "[SIT] <couch>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[FIND] <bookmark>", "[GRAB] <bookmark>", "[DROP] <bookmark>", "[TURNTO] <novel>", "[LOOKAT] <novel>", "[READ] <novel>", "[READ] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'walllamp', 'bathroom', 'orchid', 'wallshelf', 'bathroom_counter', 'floor', 'phone', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'tvstand', 'faucet', 'plate', 'mousepad', 'wall', 'powersocket', 'bench', 'oven', 'cutting_board', 'tape', 'cup', 'doorjamb', 'sink', 'light', 'table']\nTask:Pick up phone\nresponse:\n{\n    \"thoughts\": walk over and pick up the phone,\n    \"actions\": ['[WALK] <phone>', '[GRAB] <phone>']\n}\n", "input": "\nScene_Objects:['food_pizza', 'tablelamp', 'toilet', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'faucet', 'mousepad', 'wall', 'nightstand', 'cup', 'coffe_maker', 'doorjamb', 'mouse', 'chessboard', 'controller', 'bathroom_cabinet', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "enter office, find computer, turn it on,", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TOUCH] <keyboard>", "[TYPE] <keyboard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'filing_cabinet', 'sheets', 'detergent', 'milk', 'hanger', 'phone', 'desk', 'clothes_jacket', 'plate', 'dish_soap', 'mousepad', 'couch', 'candle', 'pillow', 'bowl', 'basket_for_clothes', 'standingmirror', 'keys', 'food_food', 'check', 'razor', 'coffee_filter', 'mouse', 'kitchen_counter', 'controller', 'door', 'bathtub', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Put dishes in the dishwasher and start the dishwasher.,\n    \"actions\": ['[RUN] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[WALK] <dishwasher>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'computer', 'button', 'mat', 'balanceball', 'window', 'candle', 'wall', 'powersocket', 'photoframe', 'novel', 'coffee_filter', 'ice', 'doorjamb', 'kettle', 'console', 'door', 'bathroom_cabinet', 'light', 'home_office', 'table', 'soap']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk to wall. I put my hand on the switch. I push the button. I turn on the lights.", "actions": ["[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'drawing', 'orchid', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'bench', 'photoframe', 'check', 'television', 'curtain', 'doorjamb', 'laundry_detergent', 'door', 'table', 'light', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn tv on, watch TV,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[TURNTO] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['food_bread', 'scissors', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'knifeblock', 'orchid', 'hanger', 'floor', 'bedroom', 'tea_bag', 'envelope', 'clothes_jacket', 'toaster', 'balanceball', 'faucet', 'iron', 'bed', 'coffe_maker', 'wooden_spoon', 'cards', 'headset', 'dresser', 'napkin', 'spoon', 'vase', 'instrument_violin', 'laser_pointer', 'creditcard']\nTask:Receive credit card\nresponse:\n", "output": {"thoughts": "hear the dooble, go to door, take de envelope, open it with a scissor, take the creditcard", "actions": ["[WALK] <home_office>", "[WALK] <envelope>", "[FIND] <envelope>", "[GRAB] <envelope>", "[WALK] <home_office>", "[WALK] <scissors>", "[FIND] <scissors>", "[OPEN] <envelope>", "[FIND] <creditcard>", "[GRAB] <creditcard>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'beer', 'detergent', 'food_butter', 'orchid', 'hanger', 'floor', 'ceiling', 'computer', 'toaster', 'mat', 'closetdrawer', 'broom', 'wall', 'nightstand', 'toy', 'newspaper', 'toilet_paper', 'television', 'wine', 'doorjamb', 'kitchen_counter', 'shoes', 'chessboard', 'dresser', 'door', 'dishwasher', 'instrument_violin', 'laser_pointer']\nTask:Put away shoes\nresponse:\n{\n    \"thoughts\": Pick up shoes from entry hall. Open entry way closet. Put shoes in closet. Close closet door.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[FIND] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['tea', 'jelly', 'dog', 'walllamp', 'bathroom_counter', 'thread', 'table_cloth', 'tvstand', 'balanceball', 'shower', 'window', 'oil', 'microphone', 'clothes_hat', 'nightstand', 'photoframe', 'bed', 'check', 'coffe_maker', 'toothbrush_holder', 'kitchen_counter', 'dresser', 'cupboard', 'controller', 'laundry_detergent', 'maindoor', 'sink', 'bathroom_cabinet', 'vase', 'soap']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "take remote control, push \"change channel\" button", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['crayon', 'needle', 'clothes_underwear', 'phone', 'desk', 'stamp', 'clothes_jacket', 'stovefan', 'mat', 'tvstand', 'window', 'towel', 'pillow', 'cutting_board', 'basket_for_clothes', 'cup', 'bed', 'photoframe', 'newspaper', 'food_food', 'mouthwash', 'television', 'wine', 'kettle', 'dresser', 'bathtub', 'bathroom_cabinet', 'table', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to the kitchen and find the basket of clothes. Put the soap and clothes into the washing machine. Turn on the washing machine.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[FIND] <washing_machine>', '[TURNTO] <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[OPEN] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <clothes_jacket>', '[GRAB] <clothes_jacket>', '[PUTBACK] <clothes_jacket> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'pot', 'food_pizza', 'food_ice_cream', 'freezer', 'walllamp', 'knife', 'bathroom_counter', 'floor', 'stamp', 'tea_bag', 'keyboard', 'food_cake', 'bookmark', 'toaster', 'plate', 'stereo', 'dish_soap', 'food_noodles', 'wall', 'oven', 'bench', 'coffe_maker', 'food_food', 'tray', 'food_onion', 'dishrack', 'conditioner', 'bathroom_cabinet', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find dishwasher. Open dishwasher. Find dishrack. Find plate 1. Grab plate 1. Put plate 1 in dishrack. Repeat for dish 2. Find dish soap. Grab dish soap. Put dish soap to dishwasher. Close dishwasher. Switch on dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishrack>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'toilet', 'freezer', 'cat', 'orchid', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'oil', 'couch', 'clothes_hat', 'wall', 'basket_for_clothes', 'food_food', 'coffe_maker', 'curtain', 'doorjamb', 'towel_rack', 'comb', 'dresser', 'maindoor', 'soap']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I put my groceries into the fridge.,\n    \"actions\": ['[WALK] <food_food>', '[GRAB] <food_food>', '[FIND] <freezer>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'toilet', 'ceilingfan', 'food_egg', 'hanger', 'ceiling', 'dvd_player', 'character', 'thread', 'clothes_jacket', 'toaster', 'deck_of_cards', 'foundation', 'mousepad', 'oil', 'pillow', 'oven', 'toy', 'food_food', 'check', 'lighter', 'centerpiece', 'dishrack', 'coffee', 'hair', 'laundry_detergent', 'maindoor', 'sink', 'door', 'teeth']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into my bedroom, go to the lamp and use the switch to turn it on.", "actions": ["[WALK] <bedroom>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[TURNTO] <floor_lamp>", "[LOOKAT] <floor_lamp>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'walllamp', 'knife', 'desk', 'bookshelf', 'button', 'clothes_jacket', 'stovefan', 'deck_of_cards', 'mousepad', 'closetdrawer', 'oil', 'towel', 'pillow', 'broom', 'basket_for_clothes', 'photoframe', 'novel', 'coffe_maker', 'television', 'wine', 'curtain', 'after_shave', 'cards', 'home_office', 'light', 'clothes_skirt', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": walk to the bathroom, grab the basket for clothes, find soap, put in washing machine, turn on,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[WALK] <washing_machine>', '[PUTBACK] <soap> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'microwave', 'chair', 'freezer', 'food_butter', 'drawing', 'food_egg', 'floor', 'ceiling', 'character', 'stovefan', 'ceilinglamp', 'faucet', 'shower', 'couch', 'wall', 'nightstand', 'bench', 'oven', 'cup', 'novel', 'ice', 'curtain', 'doorjamb', 'kettle', 'pencil', 'headset', 'sink', 'bathroom_cabinet', 'light']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Walk into kitchen, look for cup, walk to cup, look for refrigerator, walk to refrigerator, open door, grab water, pour water into cup", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[TURNTO] <cup>", "[LOOKAT] <cup>", "[WALK] <cup>", "[FIND] <freezer>", "[TURNTO] <freezer>", "[LOOKAT] <freezer>", "[WALK] <freezer>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'bathroom', 'trashcan', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'alcohol', 'mat', 'electric_shaver', 'oil', 'wall', 'powersocket', 'box', 'food_cheese', 'photoframe', 'bed', 'food_food', 'coffe_maker', 'doorjamb', 'lighter', 'dishrack', 'door', 'bathroom_cabinet', 'light', 'dishwasher', 'coffee']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": push any ligth button to turn it on, for example in the living room,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'chair', 'band-aids', 'detergent', 'drawing', 'needle', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'stovefan', 'ceilinglamp', 'tvstand', 'towel', 'dining_room', 'wall', 'basket_for_clothes', 'novel', 'after_shave', 'doorjamb', 'centerpiece', 'headset', 'dresser', 'cupboard', 'door', 'light', 'scrabble']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Switch on light in home office", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[PLUGIN] <light>", "[SWITCHON] <light>", "[WALK] <home_office>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'toilet', 'walllamp', 'food_carrot', 'knife', 'wallshelf', 'thread', 'envelope', 'computer', 'stovefan', 'ceilinglamp', 'shower', 'couch', 'pillow', 'broom', 'wall', 'bench', 'powersocket', 'tape', 'shelf', 'basket_for_clothes', 'photoframe', 'bed', 'oven_mitts', 'towel_rack', 'shaving_cream', 'kitchen_counter', 'shoes', 'table', 'shoe_rack']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk to the living room. I find a radio there and switch it on to listen to music.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['scissors', 'microwave', 'toilet', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'food_egg', 'cpuscreen', 'clothes_underwear', 'tea_bag', 'keyboard', 'faucet', 'clothes_socks', 'food_noodles', 'dining_room', 'bench', 'cup', 'bed', 'check', 'toilet_paper', 'television', 'after_shave', 'console', 'comb', 'hairbrush', 'cupboard', 'bathroom_cabinet', 'board_game', 'teeth']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to my cat, and I pet him", "actions": ["[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'clothes_underwear', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'button', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door']\nTask:Work\nresponse:\n{\n    \"thoughts\": Go to the computer, sit in the chair, pull up computer screen get the mouse and keyboard load work software,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[GRAB] <mouse>', '[FIND] <mousepad>', '[PUTBACK] <mouse> <mousepad>', '[TURNTO] <computer>', '[POINTAT] <computer>', '[PULL] <mouse>', '[FIND] <button>', '[PUSH] <button>', '[GRAB] <mouse>', '[PUTOBJBACK] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'ceilingfan', 'drawing', 'knife', 'longboard', 'floor', 'ceiling', 'keyboard', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'window', 'foundation', 'clothes_socks', 'couch', 'dining_room', 'broom', 'wall', 'cup', 'coffee_table', 'television', 'toothbrush_holder', 'curtain', 'lighter', 'mouse', 'maindoor', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Sit down, turn on the TV with the remote control that is on the coffee table.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <coffee_table>", "[TURNTO] <coffee_table>", "[LOOKAT] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[LOOKAT] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[SWITCHON] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'ceilingfan', 'drawing', 'knife', 'cpuscreen', 'floor', 'ceiling', 'desk', 'remote_control', 'button', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'food_orange', 'dining_room', 'wall', 'tape', 'photoframe', 'television', 'tray', 'curtain', 'kitchen_counter', 'vacuum_cleaner', 'sink', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['crayon', 'chair', 'ground_coffee', 'trashcan', 'dough', 'toothbrush', 'knifeblock', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'mop_bucket', 'toaster', 'stovefan', 'dining_room', 'food_kiwi', 'wall', 'nightstand', 'oven', 'bed', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'door', 'light', 'board_game', 'coffee']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I walk to the kitchen and find the phone. I pick up the phone. I hold it to my ear.", "actions": ["[WALK] <dining_room>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'detergent', 'cpuscreen', 'ceiling', 'computer', 'toaster', 'brush', 'laptop', 'faucet', 'mousepad', 'clothes_pants', 'dirt', 'powersocket', 'oven', 'cup', 'bed', 'food_food', 'television', 'toothbrush_holder', 'tray', 'curtain', 'doorjamb', 'mouse', 'chessboard', 'laundry_detergent', 'sink', 'dishwasher', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom. Open washing maschine. Grab pants and put it into washing maschine. Pour laundry detergent into washing machine and close the washing maschine. Plug in washing machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[PLUGIN] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'walllamp', 'trashcan', 'knifeblock', 'drawing', 'needle', 'floor', 'ceiling', 'clothes_jacket', 'stovefan', 'toaster', 'window', 'homework', 'oil', 'dustpan', 'bowl', 'wall', 'pasta', 'oven', 'toy', 'food_food', 'oven_mitts', 'tray', 'kitchen_counter', 'napkin', 'table', 'dishwasher', 'laser_pointer', 'teeth']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "I switch on the stove and cook my groceries there.", "actions": ["[WALK] <dining_room>", "[WALK] <oven>", "[FIND] <oven>", "[SWITCHON] <oven>", "[FIND] <bowl>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <oven>", "[FIND] <food_food>", "[SWITCHOFF] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'freezer', 'walllamp', 'drawing', 'floor', 'ceiling', 'bookshelf', 'mail', 'table_cloth', 'shower', 'foundation', 'dustpan', 'microphone', 'wall', 'powersocket', 'food_food', 'check', 'television', 'rag', 'doorjamb', 'kitchen_counter', 'shoes', 'cupboard', 'door', 'sink', 'hairdryer', 'light', 'clothes_skirt', 'scrabble', 'creditcard']\nTask:Turn off light\nresponse:\n{\n    \"thoughts\": Hit the switch.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <light>', '[SWITCHOFF] <light>']\n}\n", "input": "\nScene_Objects:['glue', 'pot', 'blow_dryer', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'stovefan', 'ceilinglamp', 'balanceball', 'stereo', 'window', 'shower', 'clothes_socks', 'couch', 'candle', 'broom', 'wall', 'wall_clock', 'bed', 'television', 'curtain', 'pencil', 'kitchen_counter', 'dresser', 'napkin', 'spoon', 'table', 'dirt', 'creditcard']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Find the sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'scissors', 'food_ice_cream', 'chair', 'orchid', 'drawing', 'bills', 'knifeblock', 'floor', 'character', 'alcohol', 'mat', 'window', 'couch', 'microphone', 'pillow', 'dining_room', 'wall', 'pasta', 'clothes_gloves', 'check', 'coffe_maker', 'curtain', 'doorjamb', 'kitchen_counter', 'cupboard', 'table', 'dishwasher', 'colander', 'coffee']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": walk to entrance hall. find a table. pull table. walk to living rook. push table.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <table>', '[FIND] <table>', '[PULL] <table>', '[WALK] <home_office>', '[WALK] <table>', '[PUSH] <table>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'bookshelf', 'computer', 'mail', 'alcohol', 'toaster', 'ceilinglamp', 'balanceball', 'mousepad', 'food_orange', 'wall', 'nightstand', 'mouthwash', 'doorjamb', 'mouse', 'hairbrush', 'light']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Turn on PC and log onto email account. Type email and send it. Switch off computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[TURNTO] <computer>", "[LOOKAT] <computer>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'chair', 'freezer', 'walllamp', 'cpuscreen', 'food_egg', 'floor', 'ceiling', 'desk', 'bedroom', 'keyboard', 'computer', 'toaster', 'mat', 'ceilinglamp', 'faucet', 'picture', 'homework', 'mousepad', 'wall', 'oven', 'toy', 'newspaper', 'check', 'television', 'mouse', 'laundry_detergent', 'conditioner', 'laser_pointer']\nTask:Push in desk chair\nresponse:\n{\n    \"thoughts\": Push the desk chair until it is under the desk.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[TOUCH] <chair>', '[PUSH] <chair>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'bathroom', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'character', 'desk', 'tea_bag', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'picture', 'window', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'cutting_board', 'novel', 'check', 'television', 'curtain', 'conditioner', 'hairdryer', 'clothes_dress', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "watching TV from sofa", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[PLUGIN] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[DROP] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'bathroom', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'novel', 'food_food', 'television', 'doorjamb', 'mouse', 'pencil', 'dresser', 'door', 'shoe_rack']\nTask:Put in chair\nresponse:\n{\n    \"thoughts\": Take a comfortable chair. Place it in the entrance hall,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[GRAB] <chair>', '[WALK] <home_office>', '[WALK] <floor>', '[FIND] <floor>', '[PUTBACK] <chair> <floor>']\n}\n", "input": "\nScene_Objects:['dog', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'faucet', 'clothes_socks', 'microphone', 'wall', 'bench', 'powersocket', 'photoframe', 'food_rice', 'food_onion', 'towel_rack', 'cd', 'hairbrush', 'controller', 'maindoor', 'sink', 'bathroom_cabinet', 'home_office', 'clothes_skirt', 'vase', 'soap', 'instrument_violin', 'laser_pointer', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to bathroom, go to sink, turn the water on via faucet, grab soap, use the soap to scrub my hands, put the soap back and wash my hands. Then I stop the water with faucet.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'filing_cabinet', 'crayon', 'walllamp', 'bathroom', 'toothbrush', 'floor', 'food_cake', 'toaster', 'plate', 'shower', 'faucet', 'electric_shaver', 'dish_soap', 'broom', 'wall', 'oven', 'bench', 'photoframe', 'light_bulb', 'novel', 'coffe_maker', 'television', 'tray', 'towel_rack', 'kettle', 'dishrack', 'table', 'dishwasher', 'tooth_paste']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Go to kitchen to put dirty dishs into dishwasher with soap to clean the dirty dishes,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'jelly', 'walllamp', 'bathroom', 'ceilingfan', 'drawing', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'balanceball', 'shoe-shine_kit', 'window', 'couch', 'food_jam', 'wall', 'novel', 'television', 'curtain', 'doorjamb', 'comb', 'clothes_dress', 'sink', 'home_office', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "first i will go to living room and then take the remote and swtitch on television after that il change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[POINTAT] <remote_control>", "[TOUCH] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'walllamp', 'orchid', 'floor', 'ceiling', 'computer', 'ceilinglamp', 'tvstand', 'window', 'closetdrawer', 'couch', 'microphone', 'form', 'pillow', 'wall', 'nightstand', 'oven', 'cup', 'coffe_maker', 'television', 'curtain', 'vacuum_cleaner', 'dishrack', 'dresser', 'hairbrush', 'door', 'table', 'soap']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk to the living room. I grab the remote control. I walk to the sofa. I sit down, and press the on button on the remote.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[WALK] <couch>', '[SIT] <couch>', '[TOUCH] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_salt', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'mousepad', 'microphone', 'dining_room', 'nightstand', 'basket_for_clothes', 'bed', 'newspaper', 'coffe_maker', 'check', 'tray', 'curtain', 'doorjamb', 'mouse', 'headset', 'laundry_detergent', 'maindoor', 'bathroom_cabinet', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to laundry room. Find laundry basket. Grab laundry basket. Find washing machine. Walk to washing machine. Open washing machine. Find detergent. Grab detergent. Put in detergent. Put in clothes. Close washing machine. Turn on machine.", "actions": ["[WALK] <bathroom>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <washing_machine>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[PUTOBJBACK] <laundry_detergent>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'tablelamp', 'microwave', 'toilet', 'chair', 'dog', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'folder', 'desk', 'bookshelf', 'keyboard', 'thread', 'computer', 'mail', 'bookmark', 'mousepad', 'wall', 'nightstand', 'photoframe', 'novel', 'toilet_paper', 'tray', 'towel_rack', 'mouse', 'door', 'light']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Walk to the living room. Find a desk and grab a book. Walk to the chair, sit in the chair and read a book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <desk>', '[FIND] <novel>', '[GRAB] <novel>', '[WALK] <chair>', '[SIT] <chair>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'food_ice_cream', 'food_carrot', 'longboard', 'ceiling', 'stamp', 'food_cake', 'stovefan', 'mat', 'oil', 'food_peanut_butter', 'dining_room', 'broom', 'pasta', 'basket_for_clothes', 'bed', 'coffe_maker', 'television', 'razor', 'tray', 'after_shave', 'mouse', 'laundry_detergent', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'dishwasher', 'laser_pointer', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'freezer', 'dog', 'chair', 'nail_polish', 'bathroom', 'orchid', 'bills', 'food_egg', 'hanger', 'character', 'bedroom', 'thread', 'toaster', 'faucet', 'mousepad', 'oil', 'nightstand', 'cup', 'photoframe', 'standingmirror', 'bed', 'video_game_controller', 'wine', 'doorjamb', 'maindoor', 'bathtub', 'vase', 'clothes_shirt', 'coffee']\nTask:Read book\nresponse:\n{\n    \"thoughts\": pick up  the book open the book and read,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_bread', 'filing_cabinet', 'mop', 'toilet', 'dog', 'freezer', 'longboard', 'floor', 'ceiling', 'purse', 'ceilinglamp', 'balanceball', 'faucet', 'foundation', 'food_jam', 'wall', 'oven', 'cup', 'bed', 'novel', 'food_food', 'coffe_maker', 'coffee_filter', 'curtain', 'console', 'pencil', 'maindoor', 'table', 'light', 'dirt']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "put away groceries in fridge. open the door, put on shelf, close the door.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'juice', 'toilet', 'chair', 'freezer', 'dough', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'computer', 'toaster', 'mat', 'pillow', 'dining_room', 'wall', 'oven', 'photoframe', 'food_food', 'coffe_maker', 'tray', 'curtain', 'centerpiece', 'doorjamb', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'bag', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": walk to kitchen, walk to fridge, look at bags, grab groceries, put groceries in fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[OPEN] <freezer>', '[PUTIN] <food_food> <freezer>']\n}\n", "input": "\nScene_Objects:['freezer', 'detergent', 'dough', 'drawing', 'bills', 'clothes_underwear', 'floor', 'bedroom', 'keyboard', 'piano_bench', 'ceilinglamp', 'balanceball', 'picture', 'window', 'foundation', 'electric_shaver', 'couch', 'dustpan', 'food_kiwi', 'wall', 'keys', 'check', 'television', 'tray', 'rag', 'shoes', 'door', 'table', 'vase', 'coffee']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Relax on the sofa in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_bread', 'freezer', 'walllamp', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'tea_bag', 'bookshelf', 'thread', 'alcohol', 'ceilinglamp', 'shoe-shine_kit', 'balanceball', 'wall', 'oven', 'food_food', 'coffe_maker', 'toilet_paper', 'check', 'food_rice', 'tray', 'towel_rack', 'pencil', 'kitchen_counter', 'napkin', 'dishwasher', 'clothes_shirt']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": Open the fridge, and put groceries in the fridge. Close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['juice', 'bills', 'hanger', 'shampoo', 'phone', 'dvd_player', 'character', 'alcohol', 'mat', 'stereo', 'clothes_socks', 'mousepad', 'wall', 'pasta', 'powersocket', 'tape', 'bed', 'novel', 'food_food', 'tray', 'coffee_pot', 'pencil', 'dishrack', 'cupboard', 'clothes_scarf', 'bathroom_cabinet', 'light', 'home_office', 'vase', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the home office. Pick up the phone.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'measuring_cup', 'toilet', 'chair', 'band-aids', 'freezer', 'walllamp', 'food_carrot', 'toothbrush', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'button', 'mat', 'tvstand', 'faucet', 'mousepad', 'food_kiwi', 'wall', 'light_bulb', 'coffe_maker', 'toothbrush_holder', 'rag', 'pencil', 'sink', 'bathroom_cabinet', 'soap', 'tooth_paste', 'teeth']\nTask:Brush teeth\nresponse:\n{\n    \"thoughts\": getting ready for bed,\n    \"actions\": ['[WALK] <bathroom_counter>', '[FIND] <toothbrush>', '[GRAB] <toothbrush>', '[PUTBACK] <toothbrush> <bathroom_counter>', '[FIND] <tooth_paste>', '[SQUEEZE] <tooth_paste>', '[GRAB] <toothbrush>', '[SCRUB] <toothbrush>', '[RINSE] <toothbrush>', '[FIND] <toothbrush_holder>', '[PUTBACK] <toothbrush> <toothbrush_holder>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'food_salt', 'detergent', 'orchid', 'bathroom_counter', 'character', 'keyboard', 'remote_control', 'bookshelf', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'faucet', 'dustpan', 'nightstand', 'bench', 'tape', 'cup', 'bed', 'coffee_table', 'wooden_spoon', 'doorjamb', 'towel_rack', 'pencil', 'kitchen_counter', 'cd', 'dresser', 'hair']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk next to the coffee table. pick up the remote control from the top of the coffee table. Point the remote to the TV. Press the blue arrow on the remote.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'orchid', 'drawing', 'wallshelf', 'cpuscreen', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'picture', 'mousepad', 'pillow', 'wall', 'nightstand', 'powersocket', 'photoframe', 'television', 'curtain', 'doorjamb', 'towel_rack', 'dresser', 'bathroom_cabinet', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Go to entrance hall, switch on TV, sit back in a chair and using remote watch your favorite program.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <chair>', '[SIT] <chair>', '[TURNTO] <television>', '[WATCH] <television>', '[SWITCHOFF] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'walllamp', 'trashcan', 'knifeblock', 'floor', 'character', 'computer', 'table_cloth', 'toaster', 'plate', 'faucet', 'dish_soap', 'wall', 'oven', 'bench', 'cup', 'coffe_maker', 'food_food', 'television', 'wine', 'tray', 'pencil', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I walk into the kitchen and go to the dishwasher. I open the dishwasher and grab the dishes. Iplace the dishes into the dishwasher then close it. I add soap then turn the dishwasher on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[OPEN] <dishwasher>", "[WALK] <kitchen_counter>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'juice', 'toilet', 'chair', 'jelly', 'cat', 'dough', 'cpuscreen', 'floor', 'desk', 'bedroom', 'keyboard', 'computer', 'laptop', 'mat', 'tvstand', 'mousepad', 'wall', 'nightstand', 'love_seat', 'cup', 'clothes_gloves', 'television', 'towel_rack', 'mouse', 'dresser', 'controller', 'bathroom_cabinet', 'electrical_outlet']\nTask:Write book\nresponse:\n{\n    \"thoughts\": Writing a book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <laptop>', '[FIND] <laptop>', '[GRAB] <laptop>', '[FIND] <electrical_outlet>', '[PLUGIN] <laptop>', '[WALK] <love_seat>', '[SIT] <love_seat>', '[SWITCHON] <laptop>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['sheets', 'tablelamp', 'measuring_cup', 'dog', 'nail_polish', 'detergent', 'toothbrush', 'drawing', 'hanger', 'desk', 'computer', 'toaster', 'tvstand', 'iron', 'picture', 'window', 'shower', 'food_kiwi', 'bench', 'cup', 'food_cheese', 'photoframe', 'check', 'coffe_maker', 'television', 'kitchen_counter', 'cupboard', 'home_office', 'bathroom_cabinet', 'soap']\nTask:Read book\nresponse:\n", "output": {"thoughts": "Read a book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[PUTOBJBACK] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'tape', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'pencil', 'comb', 'dresser', 'door', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office, walk to desk, sit in chair, turn on computer, look at computer, click mouse and type on keyboard.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <mouse>', '[TOUCH] <mouse>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'beer', 'toilet', 'band-aids', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'cpuscreen', 'hanger', 'desk', 'mail', 'computer', 'ceilinglamp', 'shoe-shine_kit', 'stereo', 'iron', 'shower', 'pillow', 'nightstand', 'toy', 'bed', 'food_food', 'oven_mitts', 'coffee_filter', 'after_shave', 'curtain', 'lighter', 'cupboard', 'maindoor']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk to the living room. I find the cat. I pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'juice', 'food_salt', 'measuring_cup', 'ceilingfan', 'orchid', 'shampoo', 'bedroom', 'desk', 'table_cloth', 'toaster', 'tvstand', 'stereo', 'plate', 'bench', 'basket_for_clothes', 'coffee_filter', 'ice', 'tray', 'pencil', 'headset', 'cupboard', 'maindoor', 'napkin', 'bag', 'light', 'soap', 'washing_machine', 'creditcard']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Get dirty clothes out of basket for clothes in bedroom, carry dirty clothes to laundry room, place dirty clothes in washing machine, add laundry detergent to washing machine, turn washing machine on, set type of wash, start washing machine,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <basket_for_clothes>', '[FIND] <basket_for_clothes>', '[OPEN] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <washing_machine>', '[FIND] <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'tablelamp', 'freezer', 'walllamp', 'trashcan', 'food_butter', 'toothbrush', 'orchid', 'longboard', 'phone', 'keyboard', 'shower', 'stereo', 'oil', 'mousepad', 'clothes_pants', 'pillow', 'dining_room', 'nightstand', 'photoframe', 'newspaper', 'food_onion', 'mouse', 'shoes', 'cd', 'cupboard', 'controller', 'dishwasher', 'dirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I go home and it's dark so I can not see anything", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[FIND] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'crayon', 'dog', 'freezer', 'knife', 'orchid', 'bathroom_counter', 'character', 'bedroom', 'faucet', 'couch', 'food_orange', 'towel', 'nightstand', 'oven', 'powersocket', 'photoframe', 'check', 'video_game_controller', 'food_donut', 'food_rice', 'food_onion', 'diary', 'dresser', 'hairbrush', 'clothes_scarf', 'napkin', 'bathroom_cabinet', 'colander']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": Walk into the living room. Walk up to the cat. Pet the cat with your hand.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'bathroom', 'detergent', 'cat', 'drawing', 'floor', 'folder', 'keyboard', 'food_cake', 'table_cloth', 'ceilinglamp', 'deck_of_cards', 'window', 'couch', 'wall', 'wall_clock', 'powersocket', 'newspaper', 'television', 'pencil', 'maindoor', 'clothes_scarf', 'sink', 'home_office', 'table', 'tooth_paste', 'laser_pointer', 'teeth', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "walk to entrance hall, look at the sofa, sit on the sofa, look at the cat, allow cat to near with you, touch the cat smoothly on it head, now pet the cat smoothly", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[TURNTO] <couch>", "[LOOKAT] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'cloth_napkin', 'walllamp', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'doorjamb', 'mouse', 'pencil', 'headset', 'dresser', 'door', 'sink']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Use a cloth to wipe away any marks on the computer screen.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <cloth_napkin>', '[GRAB] <cloth_napkin>', '[WIPE] <computer>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'envelope', 'computer', 'mail', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'check', 'food_food', 'television', 'tray', 'doorjamb', 'mouse', 'chessboard', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "Walk to home office. Walk to computer. Switch on Computer.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[SWITCHON] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'beer', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'table_cloth', 'stovefan', 'ceilinglamp', 'mat', 'balanceball', 'mousepad', 'candle', 'wall', 'nightstand', 'newspaper', 'doorjamb', 'mouse', 'light', 'clothes_skirt']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Turn on computer, pull up browser, browse the internet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'tablelamp', 'ground_coffee', 'walllamp', 'trashcan', 'knifeblock', 'longboard', 'floor', 'ceiling', 'character', 'bookmark', 'toaster', 'tvstand', 'wall', 'clothes_gloves', 'keys', 'coffe_maker', 'check', 'video_game_controller', 'coffee_filter', 'toilet_paper', 'food_onion', 'curtain', 'kitchen_counter', 'controller', 'hairdryer', 'dishwasher', 'vase', 'soap']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "You walk into the kitchen and turn to the coffee maker. You grab the coffee filter and put into the coffee machine, you also grab the ground coffee and put it into the coffee machine. You turn on the coffee machine.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[TURNTO] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'glue', 'tea', 'chair', 'freezer', 'walllamp', 'knifeblock', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'keyboard', 'bookshelf', 'food_bacon', 'tvstand', 'faucet', 'window', 'mousepad', 'wall', 'powersocket', 'photoframe', 'coffe_maker', 'toilet_paper', 'diary', 'mouse', 'chessboard', 'spoon', 'home_office', 'table', 'light']\nTask:Sit in chair\nresponse:\n{\n    \"thoughts\": I sit in a chair to rest and eat.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <chair>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>']\n}\n", "input": "\nScene_Objects:['pot', 'food_bread', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'knifeblock', 'orchid', 'kitchen_cabinet', 'floor', 'ceiling', 'food_chicken', 'desk', 'food_cake', 'stovefan', 'ceilinglamp', 'toaster', 'food_orange', 'wall', 'pasta', 'oven', 'tape', 'tray', 'wooden_spoon', 'centerpiece', 'kitchen_counter', 'cupboard', 'napkin', 'dishwasher', 'clothes_shirt']\nTask:Cook some food\nresponse:\n", "output": {"thoughts": "Walk into the kitchen. Open the fridge. Take out the chicken. Open the kitchen cabinet. Take out the frying pan. Put the chicken in the frying pan. Put the frying pan on the stove. Turn on the stove. Cook the chicken.", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[OPEN] <freezer>", "[FIND] <food_chicken>", "[GRAB] <food_chicken>", "[CLOSE] <freezer>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <pot>", "[GRAB] <pot>", "[CLOSE] <kitchen_cabinet>", "[WALK] <oven>", "[PUTBACK] <pot> <oven>", "[PUTBACK] <food_chicken> <oven>", "[SWITCHON] <oven>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'trashcan', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'thread', 'computer', 'ceilinglamp', 'tvstand', 'mousepad', 'dining_room', 'wall', 'powersocket', 'bench', 'cup', 'light_bulb', 'coffe_maker', 'toilet_paper', 'television', 'doorjamb', 'towel_rack', 'mouse', 'kitchen_counter', 'table']\nTask:Work\nresponse:\n{\n    \"thoughts\": Type on a computer in the home office.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['chair', 'drawing', 'hanger', 'floor', 'ceiling', 'phone', 'computer', 'ceilinglamp', 'stereo', 'balanceball', 'window', 'clothes_socks', 'food_noodles', 'couch', 'wall', 'oven', 'tape', 'check', 'television', 'doorjamb', 'mouse', 'pencil', 'cd', 'dresser', 'cupboard', 'door', 'clothes_scarf', 'light', 'table', 'instrument_violin']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I want to relax on the sofa to some relaxing music in the living room.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <stereo>", "[PLUGIN] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'jelly', 'trashcan', 'food_egg', 'cpuscreen', 'hanger', 'bathroom_counter', 'character', 'cleaning_solution', 'bookshelf', 'computer', 'shower', 'window', 'clothes_pants', 'dining_room', 'nightstand', 'pasta', 'oven', 'tape', 'food_cheese', 'bed', 'ice', 'coffee_pot', 'dishrack', 'sink', 'table', 'light', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'juice', 'microwave', 'freezer', 'bills', 'floor', 'ceiling', 'desk', 'tea_bag', 'computer', 'ceilinglamp', 'plate', 'window', 'mousepad', 'couch', 'dining_room', 'broom', 'wall', 'powersocket', 'novel', 'food_food', 'mouthwash', 'television', 'curtain', 'rag', 'towel_rack', 'kitchen_counter', 'comb', 'door', 'laser_pointer']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the fridge, take the groceries and put in it. Closed the fridge,", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'bathroom', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'window', 'couch', 'pillow', 'wall', 'powersocket', 'photoframe', 'television', 'curtain', 'after_shave', 'doorjamb', 'mouse', 'dresser', 'door', 'table', 'light', 'dishwasher']\nTask:Change TV channel\nresponse:\n{\n    \"thoughts\": I go in the living room. Then I find the remote, sit on the couch and turn on the television. Then I use the remote the change the TV channel.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <television>', '[SWITCHON] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'microwave', 'ceilingfan', 'food_carrot', 'drawing', 'floor', 'ceiling', 'phone', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'mat', 'oil', 'couch', 'dining_room', 'wall', 'bed', 'light_bulb', 'coffe_maker', 'television', 'curtain', 'kettle', 'door', 'napkin', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk to my living room and switch on the television, and lay on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'chair', 'freezer', 'bathroom', 'trashcan', 'orchid', 'floor', 'phone', 'character', 'keyboard', 'envelope', 'clothes_jacket', 'piano_bench', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'wall', 'nightstand', 'bench', 'wall_clock', 'television', 'kitchen_counter', 'cupboard', 'sink', 'spoon', 'bathroom_cabinet']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": I walk to the entrance of the bathroom. I turn to the door. I open the door to the bathroom. I walk inside of the bathroom. I find the toilet inside of the bathroom and walk over to it. I turn and look at the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[TURNTO] <toilet>', '[LOOKAT] <toilet>', '[WALK] <toilet>', '[TURNTO] <toilet>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'food_bread', 'tablelamp', 'scissors', 'microwave', 'chair', 'walllamp', 'hanger', 'floor', 'food_cake', 'toaster', 'plate', 'shower', 'tvstand', 'faucet', 'dish_soap', 'bowl', 'wall', 'oven', 'cup', 'basket_for_clothes', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'curtain', 'maindoor', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Using dishwasher to wash dishes", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dishwasher>", "[PUTBACK] <bowl> <dishwasher>", "[GRAB] <bowl>", "[PUTBACK] <bowl> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'phone', 'stamp', 'computer', 'brush', 'mat', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'nightstand', 'photoframe', 'novel', 'coffe_maker', 'television', 'doorjamb', 'vacuum_cleaner', 'hairbrush', 'cupboard', 'table']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I pick up a book off the self that I want to read. Then, I sit down on the couch, open the book, and begin reading it.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[FIND] <novel>', '[GRAB] <novel>', '[FIND] <couch>', '[WALK] <couch>', '[SIT] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['food_ice_cream', 'toilet', 'dog', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'longboard', 'floor', 'phone', 'character', 'purse', 'computer', 'toaster', 'plate', 'balanceball', 'stereo', 'dish_soap', 'food_kiwi', 'wall', 'oven', 'tape', 'coffe_maker', 'mouthwash', 'tray', 'door', 'sink', 'napkin', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "Walk to the kitchen. Open the dishwasher. Grab plate and put it in the dishwasher. Grab dish soap and put it in the dishwasher. Switch on the dishwasher.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'cat', 'trashcan', 'drawing', 'orchid', 'knife', 'cpuscreen', 'floor', 'ceiling', 'phone', 'bookshelf', 'keyboard', 'mail', 'toaster', 'mat', 'tvstand', 'couch', 'pillow', 'broom', 'wall', 'nightstand', 'powersocket', 'television', 'food_donut', 'doorjamb', 'door', 'bathroom_cabinet', 'light', 'table', 'dirt']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": My cat looks like it wants attention, so I will go pet it on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_salt', 'microwave', 'toilet', 'freezer', 'dough', 'drawing', 'hanger', 'bathroom_counter', 'desk', 'clothes_jacket', 'table_cloth', 'stovefan', 'plate', 'stereo', 'couch', 'food_peanut_butter', 'basket_for_clothes', 'photoframe', 'novel', 'mouthwash', 'toothbrush_holder', 'food_donut', 'lighter', 'laundry_detergent', 'conditioner', 'home_office', 'dishwasher', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to the kitchen and find the basket of clothes. Put the soap and clothes into the washing machine. Turn on the washing machine.", "actions": ["[WALK] <dining_room>", "[WALK] <basket_for_clothes>", "[FIND] <basket_for_clothes>", "[FIND] <washing_machine>", "[TURNTO] <washing_machine>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[FIND] <clothes_jacket>", "[GRAB] <clothes_jacket>", "[PUTBACK] <clothes_jacket> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'detergent', 'drawing', 'orchid', 'floor', 'character', 'keyboard', 'toaster', 'mat', 'faucet', 'stereo', 'couch', 'towel', 'wall', 'wall_clock', 'nightstand', 'cup', 'coffe_maker', 'kitchen_counter', 'shoes', 'cd', 'hairbrush', 'cupboard', 'sink', 'bathroom_cabinet', 'dishwasher', 'soap', 'scrabble', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I have had my lunch and walked towards the sink and applied soap to my hands and washed cleanly.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[WASH] <hands_both>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'microwave', 'chair', 'freezer', 'detergent', 'hanger', 'floor', 'ceiling', 'phone', 'dvd_player', 'character', 'bedroom', 'keyboard', 'computer', 'stereo', 'shower', 'towel', 'wall', 'wall_clock', 'pasta', 'nightstand', 'bed', 'doorjamb', 'dresser', 'cupboard', 'conditioner', 'door', 'clothes_scarf', 'light', 'clothes_shirt']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "approach the light switch, raise hand to the switch, flip the switch", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'glue', 'tablelamp', 'microwave', 'toilet', 'walllamp', 'knifeblock', 'floor', 'ceiling', 'phone', 'food_cake', 'table_cloth', 'plate', 'faucet', 'stereo', 'dish_soap', 'towel', 'bowl', 'wall', 'nightstand', 'oven', 'coffe_maker', 'food_food', 'centerpiece', 'kitchen_counter', 'dishrack', 'cupboard', 'sink', 'spoon']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": I enter the kitchen, walk to the sink, turn on the faucet, grab the sponge and get it wet, then apply dish soap to the sponge. I take a dish, scrub it with the sponge, rinse it, and place it on the dish rack to dry, and repeat until all the dishes are clean. I rinse the sponge, put it down, and turn off the faucet.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[TURNTO] <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <sponge>', '[FIND] <dish_soap>', '[GRAB] <sponge>', '[RINSE] <sponge>', '[GRAB] <dish_soap>', '[POUR] <dish_soap> <sponge>', '[PUTOBJBACK] <dish_soap>', '[SQUEEZE] <sponge>', '[FIND] <bowl>', '[GRAB] <bowl>', '[SCRUB] <bowl>', '[RINSE] <bowl>', '[FIND] <dishrack>', '[PUTBACK] <bowl> <dishrack>', '[FIND] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishrack>', '[FIND] <spoon>', '[GRAB] <spoon>', '[SCRUB] <spoon>', '[RINSE] <spoon>', '[PUTBACK] <spoon> <dishrack>', '[RINSE] <sponge>', '[SQUEEZE] <sponge>', '[PUTOBJBACK] <sponge>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_pizza', 'tablelamp', 'freezer', 'blow_dryer', 'trashcan', 'orchid', 'hanger', 'floor', 'computer', 'button', 'stovefan', 'ceilinglamp', 'window', 'couch', 'wall', 'nightstand', 'oven', 'powersocket', 'bed', 'novel', 'tray', 'curtain', 'after_shave', 'towel_rack', 'pencil', 'maindoor', 'sink', 'light', 'dishwasher']\nTask:Go to sleep\nresponse:\n", "output": {"thoughts": "Walk to bedroom. Walk to bed, lie in bed.", "actions": ["[WALK] <bedroom>", "[WALK] <bed>", "[LIE] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'food_salt', 'food_butter', 'food_carrot', 'drawing', 'food_egg', 'floor', 'phone', 'ceiling', 'character', 'purse', 'mat', 'plate', 'window', 'mousepad', 'closetdrawer', 'couch', 'food_orange', 'bed', 'photoframe', 'tray', 'coffee_pot', 'kitchen_counter', 'shoes', 'pajamas', 'sink', 'spoon', 'bathtub', 'table', 'vase']\nTask:Set up table\nresponse:\n{\n    \"thoughts\": Place your plate on the placemat. The forks will be placed on the left side of the plate and the knife and spoon on the right and place the spoon to the right of the knife. If you want to set out food, place them above the plate.,\n    \"actions\": ['[WALK] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[DROP] <plate>', '[FIND] <fork>', '[GRAB] <fork>', '[DROP] <fork>', '[FIND] <spoon>', '[GRAB] <spoon>', '[DROP] <spoon>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'ceilingfan', 'drawing', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'food_cake', 'thread', 'faucet', 'towel', 'dining_room', 'wall', 'nightstand', 'bench', 'photoframe', 'water', 'rag', 'lighter', 'cards', 'console', 'mouse', 'conditioner', 'sink', 'bathroom_cabinet', 'table', 'soap', 'laser_pointer', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go into the bathroom and go over to the sink. Grab soap and turn on the water. Wet your hands and rub them together so that the soap lathers. Rub the soap all over your hands thoroughly. Rinse off the soap with water and turn off the sink. Dry your hands with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'detergent', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'remote_control', 'alcohol', 'table_cloth', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'couch', 'wall', 'nightstand', 'oven', 'powersocket', 'novel', 'check', 'television', 'curtain', 'rag', 'doorjamb', 'dresser', 'table', 'clothes_skirt']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": go to living room, sit on couch, get remote, switch on tv, watch tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['beer', 'walllamp', 'ceilingfan', 'orchid', 'hanger', 'longboard', 'floor', 'ceiling', 'character', 'bedroom', 'folder', 'bookshelf', 'toaster', 'ceilinglamp', 'plate', 'dish_soap', 'candle', 'wall', 'oven', 'bench', 'box', 'cup', 'clothes_gloves', 'coffe_maker', 'tray', 'console', 'maindoor', 'door', 'napkin', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and open the dishwasher. I fill it up with dirty dishes and glasses. Then I put dish soap in the dishwasher, close it and turn it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <cup>", "[GRAB] <cup>", "[PUTBACK] <cup> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'scissors', 'chair', 'freezer', 'trashcan', 'knife', 'longboard', 'floor', 'shampoo', 'computer', 'button', 'tvstand', 'stereo', 'window', 'clothes_pants', 'pillow', 'dining_room', 'wall', 'nightstand', 'bed', 'novel', 'check', 'wine', 'curtain', 'cupboard', 'hair', 'sink', 'bathroom_cabinet', 'soap', 'creditcard']\nTask:Read book\nresponse:\n{\n    \"thoughts\": I would go lie down in my bed and open the book and start reading.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <bed>', '[FIND] <bed>', '[LIE] <bed>', '[FIND] <novel>', '[GRAB] <novel>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'knifeblock', 'drawing', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'desk', 'food_cake', 'stovefan', 'toaster', 'deck_of_cards', 'faucet', 'candle', 'pillow', 'dining_room', 'oven', 'photoframe', 'check', 'food_food', 'television', 'bathroom_cabinet', 'table', 'dishwasher', 'soap', 'board_game', 'clothes_shirt']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk to living room. Find cat. Walk to cat. Pull cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[PULL] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'filing_cabinet', 'cat', 'food_butter', 'hanger', 'floor', 'ceiling', 'food_cake', 'ceilinglamp', 'deck_of_cards', 'tvstand', 'homework', 'mousepad', 'closetdrawer', 'teeth', 'food_jam', 'nightstand', 'powersocket', 'bed', 'check', 'video_game_controller', 'mouthwash', 'towel_rack', 'mouse', 'console', 'cd', 'cupboard', 'instrument_guitar', 'coffee', 'shoe_rack']\nTask:Clean\nresponse:\n{\n    \"thoughts\": Clean the welcome mat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <vacuum_cleaner>', '[FIND] <vacuum_cleaner>', '[GRAB] <vacuum_cleaner>', '[SWITCHON] <vacuum_cleaner>', '[SWITCHOFF] <vacuum_cleaner>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'food_butter', 'drawing', 'bills', 'cpuscreen', 'longboard', 'floor', 'bedroom', 'desk', 'bookshelf', 'button', 'bookmark', 'toaster', 'ceilinglamp', 'deck_of_cards', 'iron', 'tvstand', 'pillow', 'pasta', 'bench', 'powersocket', 'bed', 'coffe_maker', 'television', 'ice', 'mouse', 'dresser', 'controller']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "walk into living room, look at the phone when it ringing, grab the phone or pickup the phone", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'chair', 'freezer', 'bathroom', 'detergent', 'ceilingfan', 'milk', 'cpuscreen', 'clothes_underwear', 'floor', 'ceiling', 'desk', 'keyboard', 'bookshelf', 'computer', 'bookmark', 'ceilinglamp', 'balanceball', 'mousepad', 'wall', 'nightstand', 'food_rice', 'mouse', 'pencil', 'pajamas', 'maindoor', 'board_game', 'teeth', 'shoe_rack']\nTask:Wipe down baseboards please\nresponse:\n{\n    \"thoughts\": Use a rag to get all the dust off the baseboards please.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <rag>', '[FIND] <rag>', '[GRAB] <rag>', '[WALK] <wall>', '[PUTOBJBACK] <rag>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'stereo', 'mousepad', 'food_noodles', 'form', 'food_kiwi', 'wall', 'nightstand', 'photoframe', 'check', 'after_shave', 'doorjamb', 'towel_rack', 'mouse', 'light']\nTask:Work\nresponse:\n", "output": {"thoughts": "Go to office, find computer, switch on computer, type on keyboard, use mouse, switch off computer", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <keyboard>", "[TYPE] <keyboard>", "[FIND] <mouse>", "[PUSH] <mouse>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'orchid', 'wallshelf', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'bookshelf', 'ceilinglamp', 'faucet', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'powersocket', 'bench', 'food_food', 'water', 'doorjamb', 'towel_rack', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'table', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I turn on faucet. I hold out my hands. I get my hands wet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[TURNTO] <faucet>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <hands_both>', '[POUR] <water> <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'juice', 'microwave', 'cat', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'keyboard', 'computer', 'clothes_jacket', 'button', 'ceilinglamp', 'mat', 'window', 'couch', 'food_jam', 'wall', 'nightstand', 'cutting_board', 'cup', 'television', 'mouse', 'pencil', 'kitchen_counter', 'cupboard', 'laundry_detergent', 'table', 'vase', 'coffee']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Walk into bedrrom and to the sofa. Sit on the sofa and touch cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'phone', 'ceiling', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'picture', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'mouse', 'dresser', 'door', 'light']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": I walk into my home office and walk up to my desk. I pull out the chair and sit down. I turn on the computer. After the computer is turned on I click on the google icon. I enter Disney Cruise Line into the search bar and browse the internet.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <chair>', '[PULL] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[TURNTO] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TURNTO] <keyboard>', '[LOOKAT] <keyboard>', '[TYPE] <keyboard>']\n}\n", "input": "\nScene_Objects:['fork', 'sponge', 'scissors', 'microwave', 'ground_coffee', 'dough', 'orchid', 'needle', 'cpuscreen', 'floor', 'ceiling', 'tea_bag', 'food_cake', 'toaster', 'stovefan', 'tvstand', 'food_orange', 'pillow', 'dining_room', 'wall', 'nightstand', 'bench', 'coffe_maker', 'coffee_filter', 'hairbrush', 'napkin', 'bathroom_cabinet', 'home_office', 'dishwasher', 'vase']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open the coffee maker and put a filer, grab ground coffee and put it in the coffee maker, then close the coffee maker and switch it on", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'tablelamp', 'blender', 'food_ice_cream', 'toothbrush', 'knifeblock', 'hanger', 'dvd_player', 'alcohol', 'button', 'stovefan', 'stereo', 'homework', 'mousepad', 'powersocket', 'cup', 'food_cheese', 'bed', 'food_donut', 'tray', 'after_shave', 'doorjamb', 'lighter', 'dresser', 'home_office', 'light', 'bathroom_cabinet', 'dishwasher', 'soap', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TURNTO] <cat>', '[LOOKAT] <cat>', '[FIND] <cat>', '[WALK] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'tea', 'chair', 'freezer', 'bathroom', 'trashcan', 'toothbrush', 'cpuscreen', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'character', 'bedroom', 'computer', 'plate', 'stereo', 'window', 'food_kiwi', 'wall', 'nightstand', 'box', 'bed', 'clothes_gloves', 'coffe_maker', 'wooden_spoon', 'curtain', 'shoes', 'dishwasher', 'creditcard']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go into my bedroom and see my stereo. I decide to play music. I walk over to the stereo and turn it on. I sit down on the bed to listen to the music.", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[TURNTO] <stereo>", "[LOOKAT] <stereo>", "[SWITCHON] <stereo>", "[FIND] <bed>", "[SIT] <bed>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'walllamp', 'bathroom', 'drawing', 'floor', 'character', 'bedroom', 'keyboard', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'pasta', 'bench', 'cup', 'bed', 'television', 'curtain', 'doorjamb', 'towel_rack', 'mouse', 'headset', 'cupboard', 'instrument_guitar', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk into the living room. I bend down. I sit on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['pot', 'microwave', 'ground_coffee', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'button', 'toaster', 'tvstand', 'shower', 'picture', 'couch', 'food_jam', 'food_orange', 'dining_room', 'wall', 'powersocket', 'coffe_maker', 'toilet_paper', 'coffee_filter', 'water', 'centerpiece', 'cupboard', 'door', 'sink', 'bag', 'table', 'dishwasher', 'tooth_paste']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Walk to kitchen. Find coffee maker. Find coffee filter. Put it in the coffee maker. Find ground coffee. Put it in the coffee maker. Find Water. Put it in the coffe maker. Switch on the coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[OPEN] <coffe_maker>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[FIND] <water>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'microwave', 'toilet', 'food_carrot', 'drawing', 'orchid', 'wallshelf', 'floor', 'ceiling', 'folder', 'tea_bag', 'mat', 'shower', 'window', 'mousepad', 'couch', 'pillow', 'wall', 'pasta', 'nightstand', 'photoframe', 'novel', 'food_food', 'tray', 'curtain', 'doorjamb', 'door', 'table', 'light', 'board_game']\nTask:Read book\nresponse:\n{\n    \"thoughts\": Bring book, lie on the sofa, read book,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <novel>', '[GRAB] <novel>', '[WALK] <couch>', '[LIE] <couch>', '[READ] <novel>']\n}\n", "input": "\nScene_Objects:['glue', 'food_pizza', 'filing_cabinet', 'juice', 'toilet', 'freezer', 'trashcan', 'toothbrush', 'character', 'remote_control', 'tvstand', 'faucet', 'picture', 'deck_of_cards', 'clothes_socks', 'clothes_hat', 'pillow', 'oven', 'coffe_maker', 'coin', 'curtain', 'rag', 'towel_rack', 'dishrack', 'dresser', 'cupboard', 'laundry_detergent', 'light', 'clothes_skirt', 'teeth']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "I walk to the living room and sit on the couch.  I grab the TV remote.  I press the power button on the TV remote and the TV turns on.  I grab the surround sound remote.  I press the power button on the surround sound remote and the system turns on.", "actions": ["[WALK] <home_office>", "[WALK] <remote_control>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[PUSH] <remote_control>", "[FIND] <remote_control>", "[PUSH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'trashcan', 'drawing', 'orchid', 'floor', 'ceiling', 'character', 'computer', 'food_bacon', 'stovefan', 'mat', 'tvstand', 'window', 'mousepad', 'form', 'dining_room', 'wall', 'cutting_board', 'bed', 'clothes_gloves', 'check', 'television', 'oven_mitts', 'curtain', 'mouse', 'hair', 'napkin', 'vase', 'instrument_violin', 'laser_pointer']\nTask:Gaze out window\nresponse:\n{\n    \"thoughts\": Look at the yard from the window,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <window>', '[FIND] <window>', '[TURNTO] <window>', '[LOOKAT] <window>']\n}\n", "input": "\nScene_Objects:['microwave', 'walllamp', 'cpuscreen', 'clothes_underwear', 'hanger', 'phone', 'bedroom', 'bookshelf', 'electric_shaver', 'window', 'towel', 'wall', 'nightstand', 'oven', 'powersocket', 'basket_for_clothes', 'newspaper', 'check', 'oven_mitts', 'after_shave', 'towel_rack', 'mouse', 'kettle', 'dresser', 'laundry_detergent', 'bag', 'table', 'dishwasher', 'washing_machine', 'shoe_rack']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Pick up my dirty clothes, load the washer, add detergent, turn on washer", "actions": ["[WALK] <bedroom>", "[WALK] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[WALK] <home_office>", "[WALK] <washing_machine>", "[OPEN] <washing_machine>", "[PUTBACK] <basket_for_clothes> <washing_machine>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[PUTBACK] <laundry_detergent> <washing_machine>", "[RUN] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'nail_polish', 'trashcan', 'milk', 'drawing', 'wallshelf', 'bathroom_counter', 'ceiling', 'phone', 'bedroom', 'bookshelf', 'computer', 'faucet', 'pillow', 'wall', 'photoframe', 'toilet_paper', 'check', 'coffe_maker', 'wine', 'after_shave', 'towel_rack', 'comb', 'controller', 'sink', 'bathroom_cabinet', 'light', 'table', 'instrument_violin', 'teeth']\nTask:Get toilet paper\nresponse:\n{\n    \"thoughts\": There are extra toilet papers in the bottom cabinet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <bathroom_cabinet>', '[FIND] <bathroom_cabinet>', '[OPEN] <bathroom_cabinet>', '[FIND] <toilet_paper>', '[GRAB] <toilet_paper>', '[CLOSE] <bathroom_cabinet>']\n}\n", "input": "\nScene_Objects:['sponge', 'beer', 'microwave', 'chair', 'ceilingfan', 'trashcan', 'drawing', 'orchid', 'hanger', 'floor', 'ceiling', 'shampoo', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'dustpan', 'dining_room', 'wall', 'pasta', 'oven', 'photoframe', 'television', 'curtain', 'coffee_pot', 'comb', 'controller', 'table']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "walk to living room ,find television ,find remote control ,switch on remote control and change the channel", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[SWITCHON] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'pot', 'food_pizza', 'scissors', 'tablelamp', 'microwave', 'band-aids', 'floor', 'folder', 'food_cake', 'mail', 'computer', 'table_cloth', 'toaster', 'brush', 'plate', 'shower', 'dish_soap', 'wall', 'coffe_maker', 'oven_mitts', 'food_donut', 'tray', 'curtain', 'doorjamb', 'kitchen_counter', 'dishrack', 'cupboard', 'home_office', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen. I search and find dishwasher. I search dishes. I put dishes inside the dishwasher. I put dish soap in dishwasher. I switch on dishwasher and wait for dishes to be washed.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[OPEN] <dishwasher>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['tea', 'toilet', 'chair', 'detergent', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'character', 'bookshelf', 'purse', 'envelope', 'toaster', 'tvstand', 'shower', 'window', 'pillow', 'tape', 'cup', 'coffe_maker', 'check', 'oven_mitts', 'curtain', 'rag', 'lighter', 'mouse', 'pencil', 'hairbrush', 'table', 'vase']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "find the cat, go near the cat and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'beer', 'microwave', 'ceilingfan', 'food_carrot', 'drawing', 'floor', 'ceiling', 'phone', 'bedroom', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'plate', 'window', 'mat', 'oil', 'couch', 'dining_room', 'wall', 'bed', 'light_bulb', 'coffe_maker', 'television', 'curtain', 'kettle', 'door', 'napkin', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to my living room and switch on the television, and lay on the sofa.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['walllamp', 'ceilingfan', 'drawing', 'needle', 'cpuscreen', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'bookshelf', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'homework', 'clothes_socks', 'couch', 'pillow', 'wall', 'powersocket', 'food_cheese', 'bed', 'television', 'wooden_spoon', 'curtain', 'doorjamb', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room. Sit down on the sofa. Pick up the remote control. Turn on the television with the remote control. Put down the remote control. Watch the television.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[SWITCHON] <television>", "[PUTOBJBACK] <remote_control>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'blow_dryer', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'desk', 'keyboard', 'thread', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'couch', 'pillow', 'dining_room', 'wall', 'photoframe', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'sink', 'light', 'table', 'creditcard']\nTask:Browse internet\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on Computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'nail_polish', 'bathroom', 'cat', 'orchid', 'cpuscreen', 'longboard', 'bedroom', 'keyboard', 'clothes_jacket', 'bookmark', 'toaster', 'balanceball', 'stereo', 'mousepad', 'clothes_pants', 'pillow', 'nightstand', 'wall', 'photoframe', 'food_food', 'mouthwash', 'tray', 'diary', 'comb', 'pajamas', 'maindoor', 'bag', 'spoon', 'light']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go to the living room where the cat sleeps. Find cat. Proceed to petting cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'toilet', 'freezer', 'bathroom', 'cat', 'cpuscreen', 'bathroom_counter', 'bedroom', 'food_cake', 'computer', 'alcohol', 'toaster', 'stovefan', 'ceilinglamp', 'mousepad', 'closetdrawer', 'clothes_hat', 'nightstand', 'wall_clock', 'newspaper', 'coffee_filter', 'oven_mitts', 'kettle', 'kitchen_counter', 'dresser', 'hair', 'sink', 'napkin', 'bag', 'light']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find the cat, go near the cat and pet the cat,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['sheets', 'filing_cabinet', 'board_game', 'freezer', 'food_butter', 'toothbrush', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'bedroom', 'keyboard', 'faucet', 'stereo', 'window', 'food_peanut_butter', 'towel', 'dining_room', 'wall', 'bench', 'photoframe', 'towel_rack', 'kitchen_counter', 'coffee', 'sink', 'bathroom_cabinet', 'dishwasher', 'tooth_paste', 'teeth', 'creditcard']\nTask:Wash teeth\nresponse:\n", "output": {"thoughts": "I go to sleep and i eat the dinner so i have to clean my mouth", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <tooth_paste>", "[GRAB] <tooth_paste>", "[FIND] <toothbrush>", "[GRAB] <toothbrush>", "[DROP] <tooth_paste>", "[FIND] <teeth>", "[SCRUB] <teeth>", "[RINSE] <teeth>", "[FIND] <towel>", "[GRAB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['glue', 'shoe_rack', 'food_salt', 'beer', 'measuring_cup', 'microwave', 'trashcan', 'food_carrot', 'orchid', 'drawing', 'bookshelf', 'envelope', 'alcohol', 'ceilinglamp', 'plate', 'food_jam', 'pillow', 'oven', 'newspaper', 'photoframe', 'coffe_maker', 'food_food', 'tray', 'lighter', 'doorjamb', 'coffee_pot', 'towel_rack', 'cupboard', 'hairdryer', 'creditcard']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": find a cat, touch it gently,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <cat>', '[FIND] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['chair', 'freezer', 'trashcan', 'ceilingfan', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'mat', 'oil', 'food_orange', 'pillow', 'bowl', 'wall', 'wall_clock', 'bench', 'keys', 'light_bulb', 'food_food', 'check', 'television', 'drinking_glass', 'water', 'centerpiece', 'dishrack', 'hairbrush']\nTask:Drink\nresponse:\n", "output": {"thoughts": "go to the kitchen", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <drinking_glass>", "[GRAB] <drinking_glass>", "[FIND] <water>", "[DRINK] <drinking_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['dry_pasta', 'filing_cabinet', 'mop', 'microwave', 'trashcan', 'drawing', 'ceiling', 'character', 'bedroom', 'bookshelf', 'computer', 'stovefan', 'tvstand', 'shower', 'faucet', 'mousepad', 'closetdrawer', 'form', 'pillow', 'nightstand', 'newspaper', 'bed', 'curtain', 'towel_rack', 'dishrack', 'home_office', 'light', 'dishwasher', 'table', 'scrabble']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[FIND] <dishrack>', '[FIND] <dishrack>', '[GRAB] <dishrack>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[PUTBACK] <dishrack> <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['food_salt', 'microwave', 'trashcan', 'drawing', 'floor', 'dvd_player', 'desk', 'envelope', 'ceilinglamp', 'tvstand', 'mat', 'window', 'mousepad', 'couch', 'pillow', 'dining_room', 'broom', 'wall', 'nightstand', 'cup', 'check', 'television', 'coin', 'colander', 'after_shave', 'cd', 'cupboard', 'table', 'vase', 'soap']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to the living room. Find the sofa. Sit on the sofa. Close your eyes. Breath calmly. Relax", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'dresser', 'door', 'table', 'light']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I sit down on the couch, get the tv remote, click the on button, and change the channel to the show i want to watch,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[WALK] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[PUTOBJBACK] <remote_control>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'scissors', 'tablelamp', 'crayon', 'microwave', 'freezer', 'bathroom', 'trashcan', 'ceilingfan', 'knifeblock', 'clothes_underwear', 'longboard', 'shampoo', 'dvd_player', 'folder', 'mail', 'bookmark', 'toaster', 'shower', 'homework', 'couch', 'bed', 'clothes_gloves', 'towel_rack', 'mouse', 'pencil', 'cupboard', 'maindoor', 'table', 'laser_pointer']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I walk into the living room and I pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'filing_cabinet', 'microwave', 'chair', 'wallshelf', 'floor', 'ceiling', 'character', 'desk', 'bookshelf', 'mail', 'computer', 'toaster', 'mat', 'stovefan', 'faucet', 'mousepad', 'dining_room', 'wall', 'nightstand', 'coffe_maker', 'water', 'food_rice', 'drinking_glass', 'curtain', 'diary', 'kitchen_counter', 'cupboard', 'sink', 'dishwasher']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and approach the sink. I take a drinking glass and set it under the faucet. I turn the faucet until the water fills the glass, then turn it back. Finally I lift the glass out of the sink and drink the water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <drinking_glass>', '[GRAB] <drinking_glass>', '[PUTBACK] <drinking_glass> <sink>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[SWITCHOFF] <faucet>', '[GRAB] <drinking_glass>', '[FIND] <water>', '[DRINK] <drinking_glass>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'chair', 'ground_coffee', 'orchid', 'knifeblock', 'longboard', 'floor', 'ceiling', 'stamp', 'computer', 'toaster', 'picture', 'wall', 'nightstand', 'oven', 'powersocket', 'cup', 'bed', 'novel', 'coffe_maker', 'coffee_filter', 'water', 'kitchen_counter', 'door', 'sink', 'bathroom_cabinet', 'dishwasher', 'vase', 'colander']\nTask:Make coffee\nresponse:\n", "output": {"thoughts": "Open coffee maker. Put filter in main basket, put ground coffee in filter, fill reservoir with water, turn on coffee maker.", "actions": ["[WALK] <dining_room>", "[WALK] <coffe_maker>", "[FIND] <coffe_maker>", "[OPEN] <coffe_maker>", "[FIND] <coffee_filter>", "[GRAB] <coffee_filter>", "[PUTBACK] <coffee_filter> <coffe_maker>", "[FIND] <ground_coffee>", "[GRAB] <ground_coffee>", "[FIND] <ground_coffee>", "[PUTBACK] <ground_coffee> <coffe_maker>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <coffe_maker>", "[CLOSE] <coffe_maker>", "[SWITCHON] <coffe_maker>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['toilet', 'walllamp', 'bathroom', 'drawing', 'cpuscreen', 'floor', 'dvd_player', 'desk', 'bedroom', 'bookshelf', 'computer', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'television', 'tray', 'curtain', 'controller', 'sink', 'bathroom_cabinet', 'table', 'home_office', 'light']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dvd_player>', '[FIND] <dvd_player>', '[TURNTO] <dvd_player>', '[LOOKAT] <dvd_player>', '[SWITCHON] <dvd_player>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['sponge', 'food_salt', 'chair', 'detergent', 'trashcan', 'dough', 'drawing', 'character', 'computer', 'button', 'brush', 'mat', 'couch', 'clothes_pants', 'clothes_hat', 'wall', 'nightstand', 'basket_for_clothes', 'food_cheese', 'rag', 'kitchen_counter', 'shoes', 'vacuum_cleaner', 'laundry_detergent', 'conditioner', 'sink', 'home_office', 'table', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.", "actions": ["[WALK] <bathroom>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[OPEN] <washing_machine>", "[FIND] <basket_for_clothes>", "[GRAB] <basket_for_clothes>", "[FIND] <clothes_pants>", "[GRAB] <clothes_pants>", "[PUTBACK] <clothes_pants> <washing_machine>", "[FIND] <clothes_shirt>", "[GRAB] <clothes_shirt>", "[PUTBACK] <clothes_shirt> <washing_machine>", "[PUTOBJBACK] <basket_for_clothes>", "[WALK] <laundry_detergent>", "[FIND] <laundry_detergent>", "[GRAB] <laundry_detergent>", "[POUR] <laundry_detergent> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['juice', 'food_salt', 'walllamp', 'detergent', 'trashcan', 'knifeblock', 'drawing', 'floor', 'stamp', 'table_cloth', 'plate', 'faucet', 'deck_of_cards', 'clothes_pants', 'wall', 'bench', 'powersocket', 'tape', 'photoframe', 'food_food', 'tray', 'pencil', 'kitchen_counter', 'cupboard', 'maindoor', 'sink', 'bathroom_cabinet', 'table', 'dishwasher', 'coffee']\nTask:Bring dirty plate to sink\nresponse:\n{\n    \"thoughts\": Stand up from the table.  Pick up dirty plate.  Walk to sink.  Put plate in sink.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <table>', '[FIND] <table>', '[TURNTO] <table>', '[FIND] <plate>', '[GRAB] <plate>', '[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[PUTBACK] <plate> <sink>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'crayon', 'bills', 'knifeblock', 'hanger', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'mat', 'form', 'wall', 'nightstand', 'coffe_maker', 'food_food', 'razor', 'coin', 'video_game_controller', 'check', 'tray', 'doorjamb', 'towel_rack', 'kitchen_counter', 'cupboard', 'spoon', 'napkin', 'dishwasher', 'vase']\nTask:Read book\nresponse:\n", "output": {"thoughts": "walk to living room ,find a book ,open the book ,read the book", "actions": ["[WALK] <home_office>", "[WALK] <novel>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'tea', 'microwave', 'jelly', 'band-aids', 'freezer', 'bathroom', 'needle', 'ceiling', 'character', 'keyboard', 'mop_bucket', 'ceilinglamp', 'mat', 'tvstand', 'faucet', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'oven', 'photoframe', 'coffe_maker', 'check', 'television', 'cards', 'laundry_detergent', 'clothes_dress', 'home_office', 'table']\nTask:Put umbrella away\nresponse:\n{\n    \"thoughts\": Place the umbrella in the umbrella bucket point first.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <mop_bucket>']\n}\n", "input": "\nScene_Objects:['fork', 'microwave', 'freezer', 'blow_dryer', 'food_carrot', 'milk', 'drawing', 'floor', 'phone', 'keyboard', 'bookshelf', 'button', 'piano_bench', 'laptop', 'plate', 'oil', 'mousepad', 'wall', 'bench', 'photoframe', 'keys', 'food_food', 'check', 'tray', 'kitchen_counter', 'dishrack', 'hair', 'table', 'soap', 'teeth']\nTask:Set up table\nresponse:\n", "output": {"thoughts": "I walk to the table. I put down the dish. I put down the silverware. I put down the napkin.", "actions": ["[WALK] <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>", "[GRAB] <plate>", "[PUTBACK] <plate> <table>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'walllamp', 'trashcan', 'drawing', 'orchid', 'knifeblock', 'floor', 'ceiling', 'bedroom', 'remote_control', 'button', 'ceilinglamp', 'tvstand', 'mat', 'faucet', 'foundation', 'dustpan', 'pillow', 'broom', 'wall', 'powersocket', 'cup', 'video_game_controller', 'television', 'doorjamb', 'hairbrush', 'door', 'sink', 'bag', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Walk to living room.  Find remote control.  Find television.  Point remote at television.  Press \"on\" button.  Watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[FIND] <remote_control>', '[TURNTO] <television>', '[POINTAT] <television>', '[FIND] <button>', '[PUSH] <button>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'jelly', 'trashcan', 'food_egg', 'cpuscreen', 'hanger', 'bathroom_counter', 'character', 'cleaning_solution', 'bookshelf', 'computer', 'shower', 'window', 'clothes_pants', 'dining_room', 'nightstand', 'pasta', 'oven', 'tape', 'food_cheese', 'bed', 'ice', 'coffee_pot', 'dishrack', 'sink', 'table', 'light', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands", "actions": ["[WALK] <bathroom>", "[WALK] <cleaning_solution>", "[FIND] <cleaning_solution>", "[SQUEEZE] <cleaning_solution>", "[FIND] <hands_both>", "[WASH] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'microwave', 'chair', 'jelly', 'ground_coffee', 'nail_polish', 'detergent', 'orchid', 'needle', 'drawing', 'hanger', 'character', 'envelope', 'ceilinglamp', 'shower', 'electric_shaver', 'tooth_paste', 'mousepad', 'food_kiwi', 'powersocket', 'oven', 'tape', 'cup', 'photoframe', 'wooden_spoon', 'cards', 'kitchen_counter', 'cd', 'instrument_violin', 'laser_pointer']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": grab the vesseles and wash,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <bowl>', '[FIND] <bowl>', '[GRAB] <bowl>', '[WALK] <dining_room>', '[WASH] <bowl>', '[WALK] <detergent>', '[FIND] <detergent>', '[RINSE] <detergent>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'food_ice_cream', 'toilet', 'freezer', 'detergent', 'ceilingfan', 'orchid', 'drawing', 'bathroom_counter', 'character', 'bookshelf', 'stovefan', 'tvstand', 'balanceball', 'electric_shaver', 'window', 'clothes_socks', 'oven', 'light_bulb', 'television', 'food_rice', 'tray', 'after_shave', 'wooden_spoon', 'towel_rack', 'kettle', 'mouse', 'dishrack', 'dresser', 'instrument_violin']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "Go into the living room and pet the cat", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'toilet', 'freezer', 'ground_coffee', 'ceilingfan', 'milk', 'floor', 'ceiling', 'shampoo', 'purse', 'ceilinglamp', 'mat', 'faucet', 'oil', 'wall', 'pasta', 'tape', 'novel', 'food_food', 'food_donut', 'wooden_spoon', 'mouse', 'kitchen_counter', 'door', 'home_office', 'bathroom_cabinet', 'vase', 'soap', 'instrument_violin', 'creditcard']\nTask:Put groceries in Fridge\nresponse:\n{\n    \"thoughts\": I walk in to the kitchen. I open the fridge. I place the groceries in the fridge. I close the fridge.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <food_food>', '[GRAB] <food_food>', '[PUTIN] <food_food> <freezer>', '[CLOSE] <freezer>']\n}\n", "input": "\nScene_Objects:['microwave', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'remote_control', 'mop_bucket', 'brush', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'form', 'food_kiwi', 'wall', 'oven', 'powersocket', 'cutting_board', 'cup', 'bed', 'toilet_paper', 'food_food', 'television', 'curtain', 'centerpiece', 'maindoor', 'napkin', 'table', 'vase']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to living room. Sit on the sofa. Grab the remote control and plug in TV.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[PLUGIN] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'bathroom_counter', 'bookshelf', 'keyboard', 'computer', 'button', 'ceilinglamp', 'tvstand', 'faucet', 'picture', 'mousepad', 'pillow', 'wall', 'basket_for_clothes', 'photoframe', 'curtain', 'shoes', 'cd', 'vacuum_cleaner', 'dresser', 'cupboard', 'clothes_dress', 'home_office', 'table', 'shoe_rack']\nTask:Take off shoes\nresponse:\n{\n    \"thoughts\": Bend down to reach your feet. Shift balance to one leg. Grab the shoe of the foot of the other leg, it is best to use one hand to grip near the front, and one towards the heel. Pull the shoe down toward the ground while also pushing on the heel away from the body. Repeat with other shoe. Place shoes in a designated area.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <shoes>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[OPEN] <dresser>', '[PUTIN] <shoes> <dresser>', '[FIND] <shoes>', '[GRAB] <shoes>', '[PULL] <shoes>', '[WALK] <dresser>', '[PUTIN] <shoes> <dresser>', '[CLOSE] <dresser>']\n}\n", "input": "\nScene_Objects:['toilet', 'walllamp', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'purse', 'stovefan', 'ceilinglamp', 'stereo', 'faucet', 'window', 'balanceball', 'electric_shaver', 'couch', 'wall', 'bench', 'food_food', 'television', 'curtain', 'diary', 'mouse', 'dresser', 'hair', 'bathroom_cabinet', 'table', 'home_office', 'vase', 'teeth', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk to the sofa and sit on it. Put your feet up on the sofa and lay down.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'freezer', 'knifeblock', 'cpuscreen', 'bathroom_counter', 'shampoo', 'ceiling', 'button', 'deck_of_cards', 'shower', 'clothes_pants', 'food_kiwi', 'cup', 'bed', 'food_food', 'tray', 'wooden_spoon', 'doorjamb', 'mouse', 'kitchen_counter', 'dresser', 'cupboard', 'door', 'spoon', 'home_office', 'light', 'clothes_skirt', 'dishwasher', 'coffee']\nTask:Grab things\nresponse:\n{\n    \"thoughts\": The robot will need some kind of hands or grabbing device to transport objects.,\n    \"actions\": ['[WALK] <novel>', '[TURNTO] <novel>', '[GRAB] <novel>']\n}\n", "input": "\nScene_Objects:['pot', 'filing_cabinet', 'tablelamp', 'scissors', 'freezer', 'ceilingfan', 'knifeblock', 'knife', 'hanger', 'phone', 'desk', 'clothes_jacket', 'toaster', 'ceilinglamp', 'electric_shaver', 'window', 'mousepad', 'couch', 'candle', 'cutting_board', 'food_cheese', 'food_rice', 'food_onion', 'pencil', 'cd', 'comb', 'door', 'maindoor', 'napkin', 'soap']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "You walk to the living room and find the cat. You look at the cat and touch it to pet it.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'beer', 'toilet', 'chair', 'ceilingfan', 'drawing', 'cpuscreen', 'longboard', 'floor', 'ceiling', 'folder', 'desk', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'couch', 'microphone', 'wall', 'bench', 'light_bulb', 'television', 'curtain', 'food_steak', 'sink', 'napkin', 'home_office', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into my living room, find the remote and sit on my sofa. I look at the remote and switch on the television. I look at the Television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'blender', 'knife', 'orchid', 'drawing', 'longboard', 'character', 'desk', 'computer', 'button', 'toaster', 'tvstand', 'faucet', 'clothes_socks', 'dirt', 'cutting_board', 'basket_for_clothes', 'cup', 'photoframe', 'bed', 'novel', 'coffe_maker', 'razor', 'curtain', 'mouse', 'pencil', 'table', 'dishwasher', 'soap', 'washing_machine']\nTask:Wash clothes\nresponse:\n", "output": {"thoughts": "put clothes and soap in the washing machine, turn it on", "actions": ["[WALK] <dining_room>", "[WALK] <washing_machine>", "[FIND] <washing_machine>", "[FIND] <basket_for_clothes>", "[OPEN] <basket_for_clothes>", "[FIND] <soap>", "[GRAB] <soap>", "[OPEN] <washing_machine>", "[PUTBACK] <soap> <washing_machine>", "[CLOSE] <washing_machine>", "[SWITCHON] <washing_machine>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'chair', 'walllamp', 'blow_dryer', 'orchid', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'mat', 'ceilinglamp', 'tvstand', 'mousepad', 'closetdrawer', 'pillow', 'wall', 'television', 'ice', 'rag', 'doorjamb', 'mouse', 'controller', 'door', 'light', 'table', 'scrabble', 'dirt']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to home office ,find computer ,switch on computer and work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['drawing', 'bathroom_counter', 'floor', 'ceiling', 'tea_bag', 'computer', 'button', 'mat', 'ceilinglamp', 'shower', 'faucet', 'tvstand', 'food_peanut_butter', 'towel', 'wall', 'pasta', 'photoframe', 'coffe_maker', 'curtain', 'mouse', 'kitchen_counter', 'shoes', 'laundry_detergent', 'food_steak', 'sink', 'bag', 'bathroom_cabinet', 'door', 'vase', 'soap']\nTask:Take shower\nresponse:\n", "output": {"thoughts": "I want to take a shower so I go to the bathroom, turn on the shower, get the soap, lather and wash, then I rinse, turn off shower, get a towel and dry off.", "actions": ["[WALK] <bathroom>", "[WALK] <shower>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[RINSE] <shower>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_salt', 'chair', 'walllamp', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'remote_control', 'computer', 'stovefan', 'mat', 'tvstand', 'ceilinglamp', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'food_food', 'television', 'coin', 'tray', 'doorjamb', 'towel_rack', 'dresser', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Watch the television,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[TURNTO] <television>', '[LOOKAT] <television>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[SWITCHON] <television>', '[FIND] <couch>', '[TURNTO] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'chair', 'walllamp', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'bedroom', 'desk', 'computer', 'ceilinglamp', 'deck_of_cards', 'stereo', 'window', 'couch', 'clothes_pants', 'wall', 'cup', 'novel', 'toilet_paper', 'television', 'wine', 'kitchen_counter', 'dishrack', 'comb', 'dresser', 'coffee_cup', 'table', 'teeth']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "you walk to the living room, open the door, find a comfy sofa, sit on it, read a book or sip a cup of coffee and relax.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[SIT] <couch>", "[FIND] <novel>", "[GRAB] <novel>", "[READ] <novel>", "[DRINK] <coffee_cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'shower', 'closetdrawer', 'couch', 'food_orange', 'pillow', 'food_kiwi', 'wall', 'photoframe', 'newspaper', 'television', 'food_donut', 'food_rice', 'doorjamb', 'mouse', 'cd', 'dresser']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": turn on the tv,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <television>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['glue', 'sheets', 'tablelamp', 'scissors', 'measuring_cup', 'knifeblock', 'orchid', 'desk', 'mop_bucket', 'toaster', 'deck_of_cards', 'window', 'couch', 'towel', 'pillow', 'broom', 'nightstand', 'toothbrush_holder', 'food_onion', 'centerpiece', 'mouse', 'vacuum_cleaner', 'hairbrush', 'door', 'napkin', 'home_office', 'table', 'bag', 'light', 'teeth']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to the living room, Find Phone stand. Find Phone. Pickup the receiver to attend the call.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'dustpan', 'dining_room', 'wall', 'nightstand', 'pasta', 'tape', 'photoframe', 'tray', 'rag', 'doorjamb', 'towel_rack', 'mouse', 'bathroom_cabinet', 'light']\nTask:Turn on computer\nresponse:\n{\n    \"thoughts\": Press the power button on the computer to turn it on,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'food_salt', 'microwave', 'chair', 'freezer', 'band-aids', 'bathroom', 'detergent', 'knife', 'knifeblock', 'bathroom_counter', 'longboard', 'dvd_player', 'desk', 'bookshelf', 'computer', 'bookmark', 'stovefan', 'tvstand', 'faucet', 'balanceball', 'powersocket', 'bench', 'check', 'razor', 'after_shave', 'dresser', 'controller', 'table', 'scrabble']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk to home office 1. Walk over to phone 1 on table 1.  Grab phone 1.  lift phone 1.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'beer', 'measuring_cup', 'toilet', 'drawing', 'knife', 'hanger', 'floor', 'desk', 'tea_bag', 'keyboard', 'piano_bench', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'laptop', 'foundation', 'clothes_pants', 'towel', 'wall', 'nightstand', 'novel', 'television', 'tray', 'curtain', 'bag', 'table', 'vase', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": BEFORE EATING,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <soap>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <towel>', '[WIPE] <towel>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_salt', 'tea', 'mop', 'microwave', 'toilet', 'orchid', 'longboard', 'character', 'stamp', 'purse', 'stovefan', 'laptop', 'mat', 'shower', 'faucet', 'clothes_socks', 'food_peanut_butter', 'wall', 'box', 'bed', 'coffe_maker', 'doorjamb', 'towel_rack', 'dishrack', 'maindoor', 'sink', 'dishwasher', 'clothes_shirt', 'coffee']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "I walk into the living room and then I walk to the light switch in there and turn the lights on", "actions": ["[WALK] <home_office>", "[WALK] <floor_lamp>", "[SWITCHON] <floor_lamp>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'measuring_cup', 'ground_coffee', 'walllamp', 'trashcan', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'keyboard', 'stovefan', 'toaster', 'mat', 'faucet', 'shower', 'wall', 'nightstand', 'oven', 'coffe_maker', 'coffee_filter', 'water', 'tray', 'kitchen_counter', 'cd', 'cupboard', 'controller', 'hairdryer', 'sink', 'home_office']\nTask:Make coffee\nresponse:\n{\n    \"thoughts\": Walk to Kitchen and find coffee maker, find coffee filter and place it in coffee maker. Find ground coffee and water. put both in coffee maker. Close the coffee maker and switch it on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <coffe_maker>', '[FIND] <coffe_maker>', '[OPEN] <coffe_maker>', '[FIND] <coffee_filter>', '[GRAB] <coffee_filter>', '[PUTBACK] <coffee_filter> <coffe_maker>', '[FIND] <ground_coffee>', '[GRAB] <ground_coffee>', '[PUTBACK] <ground_coffee> <coffe_maker>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <coffe_maker>', '[CLOSE] <coffe_maker>', '[SWITCHON] <coffe_maker>']\n}\n", "input": "\nScene_Objects:['pot', 'tablelamp', 'cat', 'ceilingfan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'folder', 'character', 'keyboard', 'remote_control', 'button', 'table_cloth', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'oil', 'couch', 'pillow', 'wall', 'check', 'television', 'curtain', 'door', 'home_office', 'table', 'bathroom_cabinet']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Pickup remote control for the living room TV and push the change channel button", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'toilet', 'freezer', 'band-aids', 'knifeblock', 'food_egg', 'shampoo', 'phone', 'bookshelf', 'purse', 'thread', 'deck_of_cards', 'shower', 'picture', 'clothes_socks', 'oven', 'food_cheese', 'novel', 'food_food', 'check', 'coffe_maker', 'tray', 'after_shave', 'curtain', 'doorjamb', 'cupboard', 'controller', 'bag', 'table', 'scrabble']\nTask:Drink\nresponse:\n{\n    \"thoughts\": Pour water into a glass and drink from the glass.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <water_glass>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'food_ice_cream', 'mop', 'nail_polish', 'detergent', 'trashcan', 'food_carrot', 'drawing', 'hanger', 'longboard', 'floor', 'keyboard', 'ceilinglamp', 'plate', 'faucet', 'window', 'couch', 'wall', 'nightstand', 'powersocket', 'bench', 'novel', 'food_food', 'television', 'wine', 'doorjamb', 'conditioner', 'bathroom_cabinet', 'table', 'clothes_skirt']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk into the living room. Walk up to the sofa. Sit down on the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'walllamp', 'bathroom', 'trashcan', 'bathroom_counter', 'floor', 'phone', 'bedroom', 'keyboard', 'bookshelf', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'shower', 'faucet', 'window', 'food_peanut_butter', 'clothes_pants', 'oven', 'basket_for_clothes', 'food_food', 'television', 'kettle', 'dresser', 'laundry_detergent', 'door', 'bathtub', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Walk to bathroom, open wash-machine, lift laundry basket, grab cloth and put into wash-machine one by one, put down laundry basket, pour detergent, close wash-machine, turn on wash-machine.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[OPEN] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[FIND] <clothes_pants>', '[GRAB] <clothes_pants>', '[PUTBACK] <clothes_pants> <washing_machine>', '[FIND] <clothes_shirt>', '[GRAB] <clothes_shirt>', '[PUTBACK] <clothes_shirt> <washing_machine>', '[PUTOBJBACK] <basket_for_clothes>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['spectacles', 'tablelamp', 'blender', 'food_ice_cream', 'toothbrush', 'knifeblock', 'hanger', 'dvd_player', 'alcohol', 'button', 'stovefan', 'stereo', 'homework', 'mousepad', 'powersocket', 'cup', 'food_cheese', 'bed', 'food_donut', 'tray', 'after_shave', 'doorjamb', 'lighter', 'dresser', 'home_office', 'light', 'bathroom_cabinet', 'dishwasher', 'soap', 'creditcard']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "I enter the room where the cat is. I look for the cat. I find the cat. I walk over to the cat. I use my hand to pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <cat>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[FIND] <cat>", "[WALK] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_bread', 'filing_cabinet', 'toilet', 'jelly', 'walllamp', 'drawing', 'orchid', 'needle', 'bathroom_counter', 'floor', 'folder', 'keyboard', 'toaster', 'ceilinglamp', 'window', 'food_noodles', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'light_bulb', 'television', 'curtain', 'food_onion', 'chessboard', 'cupboard', 'door', 'napkin', 'instrument_guitar', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": walk into living room, look at the sofa which one is available, sit on the sofa, now take relax on the sofa,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['microwave', 'freezer', 'ground_coffee', 'food_carrot', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'brush', 'ceilinglamp', 'mat', 'picture', 'clothes_socks', 'couch', 'microphone', 'food_orange', 'wall', 'powersocket', 'oven', 'tape', 'cup', 'keys', 'novel', 'food_food', 'check', 'mouthwash', 'curtain', 'pajamas', 'sink']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "take every item from groceries bag and put them in the fridge", "actions": ["[WALK] <dining_room>", "[WALK] <food_food>", "[FIND] <food_food>", "[GRAB] <food_food>", "[FIND] <freezer>", "[OPEN] <freezer>", "[PUTIN] <food_food> <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'microwave', 'jelly', 'trashcan', 'food_egg', 'cpuscreen', 'hanger', 'bathroom_counter', 'character', 'cleaning_solution', 'bookshelf', 'computer', 'shower', 'window', 'clothes_pants', 'dining_room', 'nightstand', 'pasta', 'oven', 'tape', 'food_cheese', 'bed', 'ice', 'coffee_pot', 'dishrack', 'sink', 'table', 'light', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom , find cleaning solution , squeeze cleaning solution , wash hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <cleaning_solution>', '[FIND] <cleaning_solution>', '[SQUEEZE] <cleaning_solution>', '[FIND] <hands_both>', '[WASH] <hands_both>']\n}\n", "input": "\nScene_Objects:['sheets', 'glue', 'beer', 'chair', 'walllamp', 'detergent', 'knifeblock', 'floor', 'phone', 'keyboard', 'computer', 'toaster', 'stereo', 'dish_soap', 'food_noodles', 'wall', 'oven', 'cutting_board', 'box', 'coffe_maker', 'television', 'tray', 'centerpiece', 'towel_rack', 'mouse', 'door', 'maindoor', 'bag', 'spoon', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "walk to kitchen, find the dishwasher, apply dish soap into vessels or dishwasher, wash the dishes with a dishwasher", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[OPEN] <dishwasher>", "[PUTBACK] <dish_soap> <dishwasher>", "[WASH] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['chair', 'walllamp', 'bills', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'mousepad', 'closetdrawer', 'wall', 'powersocket', 'shelf', 'cup', 'television', 'doorjamb', 'mouse', 'address_book', 'dresser', 'conditioner', 'bathtub', 'spoon', 'light', 'table', 'soap']\nTask:Work\nresponse:\n{\n    \"thoughts\": Walk to home office. Walk to computer. Switch on computer. Open the address book.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[FIND] <address_book>']\n}\n", "input": "\nScene_Objects:['bathroom', 'ceilingfan', 'food_carrot', 'drawing', 'orchid', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'computer', 'tvstand', 'faucet', 'shower', 'window', 'wall', 'photoframe', 'novel', 'toilet_paper', 'curtain', 'doorjamb', 'diary', 'towel_rack', 'clothes_dress', 'sink', 'bathroom_cabinet', 'instrument_guitar', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I walk in to the bathroom. I find the soap. I find the faucet. I turn on the faucet. I wash my hands. I put the soap back. I rinse my hands. I turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[WASH] <hands_both>", "[PUTOBJBACK] <soap>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'chair', 'freezer', 'walllamp', 'bathroom', 'orchid', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'mop_bucket', 'ceilinglamp', 'mat', 'electric_shaver', 'mousepad', 'wall', 'powersocket', 'photoframe', 'food_food', 'coffe_maker', 'mouse', 'pencil', 'controller', 'door', 'table', 'teeth']\nTask:Work\nresponse:\n{\n    \"thoughts\": walk to home office ,find computer ,switch on computer and work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[FIND] <computer>', '[SWITCHON] <computer>']\n}\n", "input": "\nScene_Objects:['toilet', 'detergent', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'remote_control', 'button', 'table_cloth', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'stereo', 'window', 'clothes_socks', 'couch', 'wall', 'basket_for_clothes', 'light_bulb', 'television', 'razor', 'curtain', 'food_onion', 'towel_rack', 'coffee_pot', 'pencil', 'kitchen_counter', 'bag', 'table']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "turn on the tv", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TOUCH] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['beer', 'crayon', 'ceilingfan', 'drawing', 'needle', 'floor', 'ceiling', 'shampoo', 'keyboard', 'remote_control', 'thread', 'food_bacon', 'button', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'wall', 'photoframe', 'bed', 'television', 'coffee_filter', 'curtain', 'mouse', 'table', 'light', 'laser_pointer']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I go into the living room. I find the remote. I turn on the tv. I sit on the couch and watch tv.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[FIND] <couch>', '[SIT] <couch>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'chair', 'ceilingfan', 'knifeblock', 'orchid', 'clothes_underwear', 'hanger', 'floor', 'ceiling', 'phone', 'character', 'desk', 'computer', 'couch', 'broom', 'wall', 'coffe_maker', 'video_game_controller', 'wine', 'doorjamb', 'towel_rack', 'cards', 'pencil', 'kitchen_counter', 'mouse', 'dresser', 'cupboard', 'door', 'light', 'board_game']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk over to light switch. Turn on lights.", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'walllamp', 'cat', 'drawing', 'orchid', 'wallshelf', 'hanger', 'floor', 'ceiling', 'bookshelf', 'keyboard', 'toaster', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'window', 'plate', 'couch', 'pillow', 'wall', 'newspaper', 'coffe_maker', 'television', 'curtain', 'doorjamb', 'kitchen_counter', 'bathtub', 'light', 'table']\nTask:Pet cat\nresponse:\n{\n    \"thoughts\": My cat looks like it wants attention, so I will go pet it on the couch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <cat>', '[WALK] <couch>', '[SIT] <couch>', '[TURNTO] <cat>', '[TOUCH] <cat>']\n}\n", "input": "\nScene_Objects:['pot', 'scissors', 'ceilingfan', 'drawing', 'hanger', 'longboard', 'floor', 'ceiling', 'remote_control', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'couch', 'food_kiwi', 'wall', 'oven', 'powersocket', 'cup', 'light_bulb', 'novel', 'television', 'tray', 'curtain', 'after_shave', 'kitchen_counter', 'dresser', 'table', 'light']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room, find couch, sit on couch, find remote control , turn on tv by pressing button", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[FIND] <television>", "[TURNTO] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'measuring_cup', 'chair', 'freezer', 'drawing', 'floor', 'ceiling', 'character', 'toaster', 'plate', 'shower', 'floor_lamp', 'window', 'couch', 'food_peanut_butter', 'pillow', 'dining_room', 'wall', 'oven', 'toilet_paper', 'curtain', 'doorjamb', 'towel_rack', 'shoes', 'cd', 'laundry_detergent', 'door', 'light', 'table']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to wall, find light switch, turn on lights,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[PLUGIN] <light>', '[FIND] <floor_lamp>', '[SWITCHON] <floor_lamp>', '[WALK] <bedroom>', '[WALK] <light>', '[SWITCHON] <light>']\n}\n", "input": "\nScene_Objects:['blender', 'chair', 'dog', 'bathroom', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'folder', 'desk', 'computer', 'ceilinglamp', 'window', 'mousepad', 'couch', 'towel', 'wall', 'cup', 'television', 'curtain', 'doorjamb', 'dresser', 'cupboard', 'door', 'sink', 'light', 'table']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "I walk into the living room and turn on the light. I walk to the couch then sit on the couch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'measuring_cup', 'microwave', 'toilet', 'dog', 'bathroom', 'knifeblock', 'bills', 'bathroom_counter', 'floor', 'table_cloth', 'toaster', 'plate', 'faucet', 'couch', 'towel', 'wall', 'bed', 'oven_mitts', 'water', 'shaving_cream', 'kitchen_counter', 'shoes', 'dresser', 'cupboard', 'laundry_detergent', 'sink', 'dishwasher', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Walk to kitchen. Find sink. Walk to sink. Find soap. Put soap on sink. Pour water into hands. Scrub soap. Wash water.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <sink>', '[FIND] <sink>', '[WALK] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[PUTBACK] <soap> <sink>', '[FIND] <water>', '[GRAB] <water>', '[FIND] <hands_both>', '[POUR] <water> <hands_both>', '[SCRUB] <soap>', '[WASH] <water>']\n}\n", "input": "\nScene_Objects:['juice', 'microwave', 'freezer', 'nail_polish', 'floor', 'ceiling', 'phone', 'desk', 'computer', 'food_bacon', 'toaster', 'ceilinglamp', 'foundation', 'mousepad', 'food_jam', 'wall', 'nightstand', 'oven', 'cup', 'bed', 'food_food', 'check', 'television', 'wine', 'curtain', 'cupboard', 'door', 'bathroom_cabinet', 'home_office', 'dishwasher']\nTask:Put groceries in Fridge\nresponse:\n", "output": {"thoughts": "Go to kitchen, open fridge, get groceries, put groceries in fridge, close fridge", "actions": ["[WALK] <dining_room>", "[WALK] <freezer>", "[FIND] <freezer>", "[OPEN] <freezer>", "[FIND] <food_food>", "[GRAB] <food_food>", "[PUTIN] <food_food> <freezer>", "[CLOSE] <freezer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'filing_cabinet', 'toilet', 'freezer', 'bathroom', 'detergent', 'food_butter', 'longboard', 'bookshelf', 'stovefan', 'shower', 'pillow', 'towel', 'dining_room', 'nightstand', 'powersocket', 'toy', 'basket_for_clothes', 'newspaper', 'light_bulb', 'photoframe', 'food_food', 'coin', 'doorjamb', 'headset', 'shoes', 'hairbrush', 'cupboard', 'napkin', 'home_office']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": I walk into my room and turn the radio on so that I can listen to music.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <stereo>', '[FIND] <stereo>', '[SWITCHON] <stereo>']\n}\n", "input": "\nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'drawing', 'floor', 'bookshelf', 'toaster', 'ceilinglamp', 'stereo', 'picture', 'window', 'couch', 'towel', 'wall', 'nightstand', 'oven', 'cup', 'television', 'tray', 'kitchen_counter', 'cd', 'dresser', 'hairbrush', 'cupboard', 'home_office', 'table', 'bathroom_cabinet', 'light']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "Walk to living room. Walk to radio. Switch on the radio. Walk to sofa and sit on it.", "actions": ["[WALK] <home_office>", "[WALK] <stereo>", "[SWITCHON] <stereo>", "[WALK] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'microwave', 'ground_coffee', 'walllamp', 'milk', 'knife', 'needle', 'wallshelf', 'hanger', 'bookshelf', 'clothes_jacket', 'bookmark', 'stovefan', 'mat', 'stereo', 'faucet', 'floor_lamp', 'oil', 'pillow', 'cup', 'curtain', 'after_shave', 'kitchen_counter', 'conditioner', 'sink', 'home_office', 'table', 'dishwasher', 'soap']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": Walk over to light switch, flip the switch to the on position,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['food_bread', 'food_ice_cream', 'mop', 'microwave', 'detergent', 'ceilingfan', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'bedroom', 'remote_control', 'stovefan', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'candle', 'wall', 'bed', 'television', 'curtain', 'dresser', 'maindoor', 'table', 'dishwasher', 'vase', 'coffee']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk to livingroom, sit on couch, grab remote and turn on tv.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[WATCH] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'tablelamp', 'food_ice_cream', 'microwave', 'walllamp', 'orchid', 'bills', 'knife', 'wallshelf', 'floor', 'ceiling', 'remote_control', 'computer', 'tvstand', 'faucet', 'shower', 'window', 'closetdrawer', 'couch', 'pillow', 'nightstand', 'wall', 'light_bulb', 'coffe_maker', 'television', 'curtain', 'lighter', 'vacuum_cleaner', 'dresser', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I grab the remote.  I find the red power circular button. I point the remote at the television.   I push button down and TV screen lights up. I sit and watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[TURNTO] <remote_control>', '[LOOKAT] <remote_control>', '[FIND] <television>', '[TURNTO] <television>', '[POINTAT] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>']\n}\n", "input": "\nScene_Objects:['fork', 'water_glass', 'pot', 'food_pizza', 'walllamp', 'trashcan', 'knife', 'knifeblock', 'floor', 'ceiling', 'character', 'bookshelf', 'computer', 'toaster', 'stovefan', 'plate', 'dish_soap', 'food_kiwi', 'wall', 'oven', 'food_cheese', 'coffe_maker', 'tray', 'kitchen_counter', 'cupboard', 'spoon', 'napkin', 'coffee_cup', 'dishwasher', 'soap']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "I go to the kitchen and put plates, cutlery and glasses into the dishwasher, put the soap, close the machine and switch it on.", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <coffee_cup>", "[GRAB] <coffee_cup>", "[PUTBACK] <coffee_cup> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <pot>", "[GRAB] <pot>", "[PUTBACK] <pot> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <fork>", "[GRAB] <fork>", "[PUTBACK] <fork> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <knife>", "[GRAB] <knife>", "[PUTBACK] <knife> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <plate>", "[GRAB] <plate>", "[PUTBACK] <plate> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <spoon>", "[GRAB] <spoon>", "[PUTBACK] <spoon> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[PUTBACK] <water_glass> <dishwasher>", "[FIND] <dish_soap>", "[GRAB] <dish_soap>", "[PUTBACK] <dish_soap> <dishwasher>", "[CLOSE] <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'juice', 'chair', 'jelly', 'nail_polish', 'cat', 'trashcan', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'toaster', 'ceilinglamp', 'faucet', 'shower', 'iron', 'clothes_pants', 'food_kiwi', 'wall', 'photoframe', 'food_food', 'water', 'wooden_spoon', 'curtain', 'doorjamb', 'sink', 'bathroom_cabinet', 'teeth']\nTask:Wash teeth\nresponse:\n{\n    \"thoughts\": Wash the teeth,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <water>', '[TURNTO] <water>', '[POINTAT] <water>', '[FIND] <teeth>', '[WASH] <teeth>', '[SWITCHOFF] <faucet>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'water_glass', 'pot', 'spectacles', 'walllamp', 'food_carrot', 'kitchen_cabinet', 'shampoo', 'longboard', 'computer', 'bookmark', 'mat', 'stereo', 'shower', 'homework', 'oil', 'clothes_socks', 'dining_room', 'wall_clock', 'oven', 'powersocket', 'water', 'food_onion', 'headset', 'cd', 'dresser', 'controller', 'sink', 'napkin', 'clothes_skirt']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Go to kitchen. Open the the cabinet. Locate the water glass. Pour water into glass. Drag the bottle close to mouth. Drink water.", "actions": ["[WALK] <dining_room>", "[WALK] <kitchen_cabinet>", "[FIND] <kitchen_cabinet>", "[OPEN] <kitchen_cabinet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <water_glass>", "[POUR] <water> <water_glass>", "[CLOSE] <kitchen_cabinet>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'chair', 'freezer', 'walllamp', 'drawing', 'orchid', 'cpuscreen', 'floor', 'keyboard', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'stereo', 'window', 'couch', 'candle', 'clothes_hat', 'pillow', 'wall', 'oven', 'coffe_maker', 'television', 'curtain', 'diary', 'towel_rack', 'mouse', 'dishrack', 'home_office', 'table']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the living room. I see the couch.  I sit down on the couch, then lie down and put my feet up.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'needle', 'floor', 'ceiling', 'phone', 'remote_control', 'toaster', 'ceilinglamp', 'tvstand', 'balanceball', 'shower', 'window', 'stereo', 'iron', 'stovefan', 'couch', 'pillow', 'wall', 'television', 'food_rice', 'curtain', 'doorjamb', 'bag', 'bathroom_cabinet', 'table', 'laser_pointer']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "Walk into the living room/den. Turn on the television. Switch channels to find a program with remote. I find a program and look at the screen.", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[SWITCHON] <television>", "[FIND] <remote_control>", "[TURNTO] <remote_control>", "[TURNTO] <television>", "[LOOKAT] <television>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'chair', 'walllamp', 'ceilingfan', 'drawing', 'knifeblock', 'floor', 'ceiling', 'keyboard', 'remote_control', 'purse', 'ceilinglamp', 'tvstand', 'balanceball', 'window', 'couch', 'wall', 'oven', 'mouthwash', 'television', 'wine', 'curtain', 'after_shave', 'towel_rack', 'maindoor', 'clothes_scarf', 'hairdryer', 'table', 'light', 'dishwasher']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": SIT ON THE COUCH, TURN TV ON WITH REMOTE, FIND A SHOW,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[PUTOBJBACK] <remote_control>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['band-aids', 'ceilingfan', 'milk', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'character', 'folder', 'remote_control', 'button', 'piano_bench', 'ceilinglamp', 'tvstand', 'balanceball', 'faucet', 'window', 'foundation', 'mat', 'couch', 'dustpan', 'wall', 'nightstand', 'powersocket', 'coffee_table', 'television', 'tray', 'curtain', 'table', 'laser_pointer']\nTask:Change TV channel\nresponse:\n", "output": {"thoughts": "Go to the living room, find the remote control on the coffee table, and change the TV channel.", "actions": ["[WALK] <home_office>", "[WALK] <coffee_table>", "[FIND] <coffee_table>", "[WALK] <coffee_table>", "[TURNTO] <coffee_table>", "[LOOKAT] <coffee_table>", "[FIND] <remote_control>", "[TOUCH] <remote_control>", "[GRAB] <remote_control>", "[FIND] <television>", "[TURNTO] <television>", "[POINTAT] <television>", "[FIND] <button>", "[TOUCH] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['freezer', 'drawing', 'orchid', 'knifeblock', 'hanger', 'floor', 'ceiling', 'dvd_player', 'stamp', 'bedroom', 'desk', 'keyboard', 'toaster', 'faucet', 'shower', 'couch', 'pillow', 'wall', 'oven', 'photoframe', 'bed', 'toothbrush_holder', 'doorjamb', 'towel_rack', 'door', 'bathtub', 'home_office', 'table', 'light', 'dishwasher']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": walk to living room, look at the DVD player, switch on the dvd player, look at the sofa, sit on the sofa and listen the music,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <dvd_player>', '[FIND] <dvd_player>', '[TURNTO] <dvd_player>', '[LOOKAT] <dvd_player>', '[SWITCHON] <dvd_player>', '[FIND] <couch>', '[TURNTO] <couch>', '[LOOKAT] <couch>', '[SIT] <couch>']\n}\n", "input": "\nScene_Objects:['glue', 'walllamp', 'ceilingfan', 'drawing', 'hanger', 'floor', 'bedroom', 'computer', 'ceilinglamp', 'deck_of_cards', 'iron', 'window', 'electric_shaver', 'couch', 'food_orange', 'wall', 'wall_clock', 'bed', 'novel', 'check', 'television', 'toothbrush_holder', 'wooden_spoon', 'after_shave', 'towel_rack', 'dresser', 'conditioner', 'maindoor', 'table', 'vase']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Walk to the living room. Find the sofa. Walk to the sofa. Lie in the sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[WALK] <couch>", "[LIE] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['fork', 'pot', 'filing_cabinet', 'microwave', 'chair', 'jelly', 'bathroom', 'detergent', 'knife', 'floor', 'toaster', 'mat', 'plate', 'shower', 'faucet', 'dish_soap', 'bowl', 'wall', 'cup', 'coffe_maker', 'food_food', 'after_shave', 'kitchen_counter', 'cupboard', 'hair', 'sink', 'spoon', 'home_office', 'table', 'dishwasher']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": I walk to the kitchen and find the dishwasher. I open it. I find the sink and grab the dirty plates, cups, bowls, and utensils. I rinse off the dirty dishes under the faucet. I then place the dirty dishes into the dishwasher. I add soap to the dishwasher and close the lid. I turn it on and wait.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <sink>', '[FIND] <bowl>', '[GRAB] <bowl>', '[RINSE] <bowl>', '[PUTBACK] <bowl> <dishwasher>', '[FIND] <cup>', '[GRAB] <cup>', '[RINSE] <cup>', '[PUTBACK] <cup> <dishwasher>', '[FIND] <fork>', '[GRAB] <fork>', '[RINSE] <fork>', '[PUTBACK] <fork> <dishwasher>', '[FIND] <knife>', '[GRAB] <knife>', '[RINSE] <knife>', '[PUTBACK] <knife> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[RINSE] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <spoon>', '[GRAB] <spoon>', '[RINSE] <spoon>', '[PUTBACK] <spoon> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[TURNTO] <dishwasher>']\n}\n", "input": "\nScene_Objects:['juice', 'toilet', 'chair', 'detergent', 'bathroom_counter', 'ceiling', 'character', 'desk', 'laptop', 'tvstand', 'faucet', 'electric_shaver', 'mousepad', 'couch', 'food_orange', 'bench', 'powersocket', 'cup', 'bed', 'novel', 'razor', 'oven_mitts', 'curtain', 'after_shave', 'cd', 'headset', 'dresser', 'pajamas', 'bathroom_cabinet', 'tooth_paste']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "I'll enter the location containing the phone. Once there, I'll locate the phone, and move close to it. I'll then use my hand to pick up the phone such that I can then use it.", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[FIND] <phone>", "[WALK] <phone>", "[TURNTO] <phone>", "[LOOKAT] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'pot', 'glue', 'tablelamp', 'microwave', 'walllamp', 'knife', 'knifeblock', 'longboard', 'floor', 'phone', 'computer', 'toaster', 'plate', 'shoe-shine_kit', 'dish_soap', 'mousepad', 'food_orange', 'wall', 'nightstand', 'oven', 'toy', 'newspaper', 'coffe_maker', 'tray', 'vacuum_cleaner', 'dresser', 'hairdryer', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n{\n    \"thoughts\": Walk to the  kitchen. Open the dishwasher. Put plates into the dishwasher. Put dish soap into the dishwasher. Close the dishwasher. Turn on the dishwasher.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <dishwasher>', '[FIND] <dishwasher>', '[OPEN] <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <plate>', '[GRAB] <plate>', '[PUTBACK] <plate> <dishwasher>', '[FIND] <dish_soap>', '[GRAB] <dish_soap>', '[PUTBACK] <dish_soap> <dishwasher>', '[CLOSE] <dishwasher>', '[SWITCHON] <dishwasher>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'spectacles', 'food_pizza', 'tablelamp', 'blender', 'microwave', 'chair', 'cat', 'trashcan', 'dough', 'orchid', 'knifeblock', 'clothes_underwear', 'desk', 'food_bacon', 'shoe-shine_kit', 'balanceball', 'deck_of_cards', 'couch', 'towel', 'nightstand', 'oven', 'newspaper', 'keys', 'pencil', 'laundry_detergent', 'maindoor', 'bag', 'laser_pointer', 'shoe_rack']\nTask:Drink\nresponse:\n", "output": {"thoughts": "Drink water", "actions": ["[WALK] <dining_room>", "[WALK] <water_glass>", "[FIND] <water_glass>", "[GRAB] <water_glass>", "[DRINK] <water_glass>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'food_ice_cream', 'food_carrot', 'longboard', 'ceiling', 'stamp', 'food_cake', 'stovefan', 'mat', 'oil', 'food_peanut_butter', 'dining_room', 'broom', 'pasta', 'basket_for_clothes', 'bed', 'coffe_maker', 'television', 'razor', 'tray', 'after_shave', 'mouse', 'laundry_detergent', 'clothes_scarf', 'sink', 'bathroom_cabinet', 'dishwasher', 'laser_pointer', 'clothes_shirt', 'washing_machine']\nTask:Wash clothes\nresponse:\n{\n    \"thoughts\": Go to kitchen, find washing machine, find laundry, put laundry in washing machine, find soap, put soap in washing machine, turn on.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <washing_machine>', '[FIND] <washing_machine>', '[FIND] <basket_for_clothes>', '[GRAB] <basket_for_clothes>', '[OPEN] <washing_machine>', '[PUTBACK] <basket_for_clothes> <washing_machine>', '[WALK] <laundry_detergent>', '[FIND] <laundry_detergent>', '[GRAB] <laundry_detergent>', '[POUR] <laundry_detergent> <washing_machine>', '[CLOSE] <washing_machine>', '[SWITCHON] <washing_machine>']\n}\n", "input": "\nScene_Objects:['pot', 'food_pizza', 'tablelamp', 'walllamp', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'shampoo', 'bedroom', 'mop_bucket', 'faucet', 'balanceball', 'towel', 'wall', 'photoframe', 'keys', 'novel', 'razor', 'television', 'diary', 'kettle', 'coffee_pot', 'cd', 'sink', 'bathroom_cabinet', 'home_office', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom and locate the sink. I turn the faucet so that water is running, and rinse my hands briefly. Then I grab the soap and squeeze some onto my hands. I spend a few minutes rubbing my hands together, then I rinse them again. Finally, I turn the faucet off and wipe my hands on a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[TURNTO] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[SQUEEZE] <soap>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sheets', 'water_glass', 'food_salt', 'freezer', 'walllamp', 'blow_dryer', 'detergent', 'drawing', 'wallshelf', 'floor', 'ceiling', 'phone', 'desk', 'stovefan', 'mat', 'iron', 'oil', 'closetdrawer', 'couch', 'pillow', 'wall', 'newspaper', 'keys', 'food_food', 'television', 'kitchen_counter', 'cupboard', 'door', 'bathroom_cabinet', 'tooth_paste']\nTask:Drink\nresponse:\n{\n    \"thoughts\": walk to kitchen ,find fridge ,open fridge ,find water  glass , drink from water glass,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <freezer>', '[FIND] <freezer>', '[OPEN] <freezer>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'walllamp', 'drawing', 'knife', 'floor', 'ceiling', 'bookshelf', 'ceilinglamp', 'shower', 'food_noodles', 'oil', 'dining_room', 'wall', 'oven', 'photoframe', 'newspaper', 'check', 'coffe_maker', 'television', 'wooden_spoon', 'doorjamb', 'cards', 'pencil', 'shoes', 'cupboard', 'controller', 'door', 'maindoor', 'light', 'teeth']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "flip the switch and turn on the lights", "actions": ["[WALK] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['walllamp', 'trashcan', 'phone', 'ceiling', 'character', 'desk', 'bookmark', 'toaster', 'stovefan', 'shoe-shine_kit', 'shower', 'window', 'wall', 'pasta', 'oven', 'powersocket', 'clothes_gloves', 'novel', 'coffe_maker', 'check', 'coin', 'curtain', 'doorjamb', 'cards', 'towel_rack', 'kitchen_counter', 'chessboard', 'napkin', 'bathroom_cabinet', 'light']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": walk to living room ,find electrical outlet , switch on eletrical outlet,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <electrical_outlet>', '[FIND] <electrical_outlet>', '[SWITCHON] <electrical_outlet>']\n}\n", "input": "\nScene_Objects:['freezer', 'ceilingfan', 'drawing', 'floor', 'ceiling', 'stamp', 'remote_control', 'computer', 'clothes_jacket', 'ceilinglamp', 'tvstand', 'balanceball', 'iron', 'window', 'shoe-shine_kit', 'shower', 'mousepad', 'deck_of_cards', 'couch', 'food_kiwi', 'wall', 'cup', 'television', 'ice', 'curtain', 'conditioner', 'door', 'table', 'vase', 'tooth_paste']\nTask:Watch TV\nresponse:\n", "output": {"thoughts": "walk to living room ,find television , switch on remote control", "actions": ["[WALK] <home_office>", "[WALK] <television>", "[FIND] <television>", "[FIND] <remote_control>", "[SWITCHON] <remote_control>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['jelly', 'freezer', 'walllamp', 'bathroom', 'bills', 'needle', 'drawing', 'character', 'table_cloth', 'toaster', 'picture', 'floor_lamp', 'window', 'mousepad', 'microphone', 'food_kiwi', 'wall', 'wall_clock', 'newspaper', 'bed', 'television', 'tray', 'cupboard', 'controller', 'door', 'sink', 'bathtub', 'bathroom_cabinet', 'light', 'home_office']\nTask:Turn on light\nresponse:\n{\n    \"thoughts\": I go to bedroom and locate a lamp. This lamp can be turned on by touch, as it has capacitive sensing. So I do it.,\n    \"actions\": ['[WALK] <bedroom>', '[WALK] <floor_lamp>', '[FIND] <floor_lamp>', '[TOUCH] <floor_lamp>', '[SWITCHON] <floor_lamp>']\n}\n", "input": "\nScene_Objects:['fork', 'walllamp', 'detergent', 'orchid', 'knife', 'hanger', 'floor', 'shampoo', 'dvd_player', 'character', 'keyboard', 'mop_bucket', 'toaster', 'dining_room', 'wall', 'oven', 'cutting_board', 'light_bulb', 'coffe_maker', 'food_food', 'razor', 'tray', 'wooden_spoon', 'pencil', 'dishrack', 'cupboard', 'bathroom_cabinet', 'dishwasher', 'colander', 'laser_pointer']\nTask:Wash dishes by hand\nresponse:\n", "output": {"thoughts": "walk to kitchen , find dish washer , wash the dishe rack , put dishes in dish washer , switch on dish washer", "actions": ["[WALK] <dining_room>", "[WALK] <dishwasher>", "[FIND] <dishwasher>", "[FIND] <dishrack>", "[FIND] <dishrack>", "[GRAB] <dishrack>", "[FIND] <dishwasher>", "[OPEN] <dishwasher>", "[PUTBACK] <dishrack> <dishwasher>", "[SWITCHON] <dishwasher>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'cat', 'food_carrot', 'drawing', 'knifeblock', 'cpuscreen', 'wallshelf', 'hanger', 'bathroom_counter', 'ceiling', 'desk', 'bookshelf', 'keyboard', 'mousepad', 'bench', 'newspaper', 'coffe_maker', 'food_food', 'tray', 'curtain', 'mouse', 'door', 'hairdryer', 'sink', 'food_steak', 'light', 'dishwasher', 'spoon', 'table', 'board_game']\nTask:Write  school paper\nresponse:\n{\n    \"thoughts\": Writing  and drafting a paper for  class work,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['sponge', 'pot', 'juice', 'cd_player', 'toilet', 'freezer', 'trashcan', 'bedroom', 'remote_control', 'button', 'clothes_jacket', 'toaster', 'laptop', 'ceilinglamp', 'plate', 'homework', 'dining_room', 'broom', 'bench', 'tape', 'keys', 'coffe_maker', 'television', 'towel_rack', 'hair', 'cupboard', 'maindoor', 'sink', 'clothes_skirt', 'vase']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I walk to my living room. Find CD player. Plugin the CD player to Power. Also find the remote control. Push the Botton", "actions": ["[WALK] <home_office>", "[WALK] <cd_player>", "[FIND] <cd_player>", "[PLUGIN] <cd_player>", "[FIND] <remote_control>", "[GRAB] <remote_control>", "[FIND] <button>", "[PUSH] <button>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'pot', 'tablelamp', 'chair', 'walllamp', 'drawing', 'cpuscreen', 'wallshelf', 'floor', 'phone', 'desk', 'keyboard', 'computer', 'mat', 'deck_of_cards', 'picture', 'mousepad', 'wall', 'oven', 'powersocket', 'check', 'television', 'tray', 'doorjamb', 'towel_rack', 'cards', 'mouse', 'door', 'light', 'printing_paper']\nTask:Add paper to printer\nresponse:\n{\n    \"thoughts\": Take a stack of paper, about 25 pieces. Place carefully in the paper tray.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <desk>', '[FIND] <printing_paper>', '[GRAB] <printing_paper>', '[WALK] <fax_machine>', '[PUTBACK] <printing_paper> <fax_machine>']\n}\n", "input": "\nScene_Objects:['nail_polish', 'orchid', 'drawing', 'bathroom_counter', 'floor', 'ceiling', 'thread', 'faucet', 'couch', 'dining_room', 'wall', 'nightstand', 'bench', 'wall_clock', 'photoframe', 'coffe_maker', 'food_food', 'lighter', 'doorjamb', 'towel_rack', 'diary', 'shoes', 'comb', 'door', 'sink', 'bathroom_cabinet', 'table', 'soap', 'dirt', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom. turn on the water in the sink. apply soap to hands. rub hands together. rinse hands with water. turn off the water.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <soap>", "[GRAB] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['sponge', 'tablelamp', 'chair', 'walllamp', 'orchid', 'knifeblock', 'wallshelf', 'floor', 'ceiling', 'remote_control', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'window', 'closetdrawer', 'couch', 'pillow', 'wall', 'nightstand', 'tape', 'photoframe', 'coffee_table', 'television', 'toothbrush_holder', 'razor', 'curtain', 'dresser', 'pajamas', 'table']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": I walk into the living room and sit on the couch. I pick up the remote to turn on the television. I put down the remote on the coffee table and look at the television.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <couch>', '[FIND] <couch>', '[SIT] <couch>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[FIND] <coffee_table>', '[PUTBACK] <remote_control> <coffee_table>', '[TURNTO] <television>', '[LOOKAT] <television>']\n}\n", "input": "\nScene_Objects:['glue', 'filing_cabinet', 'tablelamp', 'microwave', 'toilet', 'dog', 'walllamp', 'bathroom', 'knifeblock', 'food_egg', 'food_bacon', 'brush', 'mat', 'deck_of_cards', 'balanceball', 'tvstand', 'ceilinglamp', 'picture', 'couch', 'wall', 'tape', 'cup', 'toilet_paper', 'food_food', 'video_game_controller', 'water', 'cupboard', 'bathroom_cabinet', 'table', 'vase']\nTask:Drink\nresponse:\n", "output": {"thoughts": "I go to the kitchen. I get a glass. I fill it with water. I take a drink from the glass.", "actions": ["[WALK] <dining_room>", "[WALK] <cup>", "[FIND] <cup>", "[GRAB] <cup>", "[WALK] <water>", "[FIND] <water>", "[GRAB] <water>", "[POUR] <water> <cup>", "[DRINK] <cup>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['filing_cabinet', 'toilet', 'chair', 'walllamp', 'trashcan', 'cpuscreen', 'floor', 'ceiling', 'character', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'faucet', 'shower', 'mousepad', 'couch', 'pillow', 'dining_room', 'wall', 'powersocket', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'table', 'coffee']\nTask:Call family member with skype application\nresponse:\n{\n    \"thoughts\": Sit down on desk chair located in front of computer.  Turn on computer.  Log in to skype application.  Locate individual to call on contacts list.  Press the call button with computer mouse.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <chair>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <computer>', '[SWITCHON] <computer>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[FIND] <mouse>', '[GRAB] <mouse>', '[TURNTO] <computer>', '[POINTAT] <computer>']\n}\n", "input": "\nScene_Objects:['sponge', 'filing_cabinet', 'tea', 'microwave', 'toilet', 'chair', 'freezer', 'walllamp', 'detergent', 'trashcan', 'toothbrush', 'orchid', 'needle', 'shampoo', 'tea_bag', 'alcohol', 'mat', 'shower', 'mousepad', 'towel', 'wall', 'cup', 'bed', 'light_bulb', 'coffe_maker', 'wooden_spoon', 'shoes', 'home_office', 'vase', 'soap']\nTask:Pick up phone\nresponse:\n", "output": {"thoughts": "Walk into the living room, walk over to phone, grab the phone, put it to your ear .", "actions": ["[WALK] <home_office>", "[WALK] <phone>", "[GRAB] <phone>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tea', 'cat', 'drawing', 'bills', 'bathroom_counter', 'floor', 'ceiling', 'character', 'bedroom', 'mail', 'alcohol', 'mat', 'faucet', 'clothes_pants', 'pillow', 'wall', 'wall_clock', 'oven', 'pasta', 'bench', 'coffee_filter', 'ice', 'paper_towel', 'conditioner', 'sink', 'bathroom_cabinet', 'clothes_skirt', 'soap', 'hands_both', 'shoe_rack']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": I walk to the bathroom and find the sink. I grab the soap and rub some on my hands. I turn the faucet on and put my hands in the water. I rinse my hands and then turn off the faucet. Then I dry my hands on a paper towel.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[RINSE] <hands_both>', '[FIND] <paper_towel>', '[GRAB] <paper_towel>', '[WIPE] <hands_both>']\n}\n", "input": "\nScene_Objects:['filing_cabinet', 'shoe_rack', 'food_salt', 'mop', 'microwave', 'dog', 'jelly', 'walllamp', 'drawing', 'needle', 'food_egg', 'desk', 'stamp', 'computer', 'ceilinglamp', 'window', 'mousepad', 'food_orange', 'pillow', 'bench', 'cup', 'oven_mitts', 'wooden_spoon', 'pencil', 'dresser', 'maindoor', 'bag', 'table', 'instrument_violin', 'creditcard']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "go to radio. run on radio.", "actions": ["[WALK] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'beer', 'crayon', 'chair', 'freezer', 'walllamp', 'orchid', 'drawing', 'hanger', 'floor', 'ceiling', 'bedroom', 'bookshelf', 'food_cake', 'ceilinglamp', 'tvstand', 'mat', 'couch', 'food_peanut_butter', 'pillow', 'wall', 'nightstand', 'powersocket', 'cup', 'food_food', 'television', 'curtain', 'pencil', 'kitchen_counter', 'door']\nTask:Relax on sofa\nresponse:\n{\n    \"thoughts\": I walk to the sofa, and lie down on the sofa.,\n    \"actions\": ['[WALK] <couch>', '[LIE] <couch>']\n}\n", "input": "\nScene_Objects:['detergent', 'trashcan', 'drawing', 'clothes_underwear', 'bathroom_counter', 'floor', 'ceiling', 'longboard', 'bookshelf', 'toaster', 'ceilinglamp', 'deck_of_cards', 'faucet', 'balanceball', 'shower', 'mousepad', 'wall', 'nightstand', 'photoframe', 'light_bulb', 'novel', 'food_food', 'water', 'kitchen_counter', 'sink', 'bag', 'bathroom_cabinet', 'napkin', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I turn on faucet. I hold out my hands. I get my hands wet.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <water>", "[GRAB] <water>", "[FIND] <hands_both>", "[POUR] <water> <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'chair', 'freezer', 'walllamp', 'trashcan', 'knifeblock', 'needle', 'cpuscreen', 'floor', 'character', 'keyboard', 'table_cloth', 'plate', 'faucet', 'window', 'oil', 'mousepad', 'pillow', 'wall', 'oven', 'food_cheese', 'bed', 'tray', 'kettle', 'kitchen_counter', 'dresser', 'cupboard', 'sink', 'home_office', 'light']\nTask:Wash dishes by hand\nresponse:\n{\n    \"thoughts\": Walking towards sink and look at the leftover plates. Scrape dishes to remove leftover food - use a rubber spatula or paper towel. Fill sink or dishpan with clean, hot water. Then wash and rinse to removing residues. Finally leave it to dry.,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <kitchen_counter>', '[FIND] <plate>', '[TURNTO] <plate>', '[LOOKAT] <plate>', '[GRAB] <plate>', '[SCRUB] <plate>', '[RINSE] <plate>', '[WASH] <plate>', '[PUTBACK] <plate> <kitchen_counter>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'bathroom', 'ceilingfan', 'knife', 'drawing', 'needle', 'knifeblock', 'bathroom_counter', 'shampoo', 'floor', 'ceiling', 'phone', 'piano_bench', 'faucet', 'window', 'teeth', 'clothes_pants', 'wall', 'photoframe', 'check', 'oven_mitts', 'tray', 'pencil', 'clothes_dress', 'sink', 'bathroom_cabinet', 'home_office', 'clothes_skirt', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "Go to the bathroom and to the sink, turn the faucet on, push the bottle of soap onto your hands and scrub them. Then rinse your hands out and turn the faucet off.", "actions": ["[WALK] <bathroom>", "[WALK] <sink>", "[FIND] <sink>", "[FIND] <faucet>", "[SWITCHON] <faucet>", "[FIND] <soap>", "[PUSH] <soap>", "[FIND] <hands_both>", "[SCRUB] <hands_both>", "[RINSE] <hands_both>", "[SWITCHOFF] <faucet>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['tablelamp', 'chair', 'walllamp', 'blow_dryer', 'orchid', 'drawing', 'cpuscreen', 'wallshelf', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'computer', 'ceilinglamp', 'tvstand', 'mat', 'mousepad', 'couch', 'pillow', 'wall', 'nightstand', 'oven', 'television', 'doorjamb', 'mouse', 'dresser', 'door', 'dishwasher']\nTask:Write an email\nresponse:\n{\n    \"thoughts\": Turn on PC and log onto email account. Type email and send it. Switch off computer.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <computer>', '[SWITCHON] <computer>', '[FIND] <keyboard>', '[TYPE] <keyboard>', '[TURNTO] <computer>', '[LOOKAT] <computer>', '[SWITCHOFF] <computer>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'ceilingfan', 'drawing', 'cpuscreen', 'phone', 'desk', 'keyboard', 'mail', 'mat', 'window', 'towel', 'dining_room', 'wall', 'pasta', 'oven', 'cup', 'check', 'curtain', 'rag', 'doorjamb', 'cards', 'cupboard', 'conditioner', 'clothes_scarf', 'napkin', 'home_office', 'vase', 'soap', 'laser_pointer']\nTask:Listen to music\nresponse:\n", "output": {"thoughts": "I go to my bedroom, and then I switch on the radio", "actions": ["[WALK] <bedroom>", "[WALK] <stereo>", "[FIND] <stereo>", "[SWITCHON] <stereo>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['spectacles', 'microwave', 'chair', 'dog', 'walllamp', 'drawing', 'cpuscreen', 'hanger', 'floor', 'ceiling', 'phone', 'desk', 'bookshelf', 'keyboard', 'computer', 'laptop', 'mat', 'mousepad', 'closetdrawer', 'couch', 'wall', 'powersocket', 'coffe_maker', 'wine', 'doorjamb', 'kettle', 'mouse', 'pencil', 'dresser', 'light']\nTask:Work\nresponse:\n{\n    \"thoughts\": I walk in. Turn on the light. Set in my chair. Turn on lap top. And start to work.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <light>', '[FIND] <light>', '[SWITCHON] <light>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <laptop>', '[GRAB] <laptop>', '[SWITCHON] <laptop>']\n}\n", "input": "\nScene_Objects:['dry_pasta', 'tablelamp', 'crayon', 'band-aids', 'nail_polish', 'cat', 'food_carrot', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'food_bacon', 'ceilinglamp', 'shoe-shine_kit', 'window', 'couch', 'food_kiwi', 'wall', 'bed', 'keys', 'check', 'television', 'oven_mitts', 'curtain', 'conditioner', 'maindoor', 'instrument_guitar', 'table', 'bathroom_cabinet']\nTask:Pet cat\nresponse:\n", "output": {"thoughts": "The cat is laying on the couch in the living room. I want to pet it, so I am going to walk to the living room, sit on the couch, and pet the cat.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[SIT] <couch>", "[FIND] <cat>", "[TURNTO] <cat>", "[LOOKAT] <cat>", "[TOUCH] <cat>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['water_glass', 'pot', 'food_salt', 'mop', 'freezer', 'walllamp', 'detergent', 'knifeblock', 'floor', 'ceiling', 'purse', 'button', 'piano_bench', 'food_bacon', 'stovefan', 'ceilinglamp', 'tvstand', 'faucet', 'form', 'wall', 'bench', 'food_food', 'television', 'water', 'kitchen_counter', 'cupboard', 'food_steak', 'sink', 'maindoor', 'table']\nTask:Drink\nresponse:\n{\n    \"thoughts\": I need to go to Kitchen.  Get a cup.  Fill cup.  Drink,\n    \"actions\": ['[WALK] <dining_room>', '[WALK] <cupboard>', '[FIND] <cupboard>', '[OPEN] <cupboard>', '[FIND] <water_glass>', '[GRAB] <water_glass>', '[WALK] <freezer>', '[WALK] <water>', '[FIND] <water>', '[GRAB] <water>', '[POUR] <water> <water_glass>', '[DRINK] <water_glass>']\n}\n", "input": "\nScene_Objects:['chair', 'food_butter', 'ceilingfan', 'drawing', 'food_egg', 'floor', 'keyboard', 'purse', 'ceilinglamp', 'window', 'couch', 'pillow', 'wall', 'nightstand', 'cup', 'food_food', 'television', 'oven_mitts', 'tray', 'curtain', 'diary', 'mouse', 'vacuum_cleaner', 'dresser', 'door', 'table', 'light', 'board_game', 'clothes_shirt', 'shoe_rack']\nTask:Relax on sofa\nresponse:\n", "output": {"thoughts": "Go to living room.  Find sofa.  Sit on sofa.", "actions": ["[WALK] <home_office>", "[WALK] <couch>", "[FIND] <couch>", "[SIT] <couch>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'food_carrot', 'orchid', 'drawing', 'wallshelf', 'bathroom_counter', 'floor', 'character', 'shoe-shine_kit', 'faucet', 'couch', 'form', 'wall', 'bench', 'toy', 'cup', 'coffe_maker', 'television', 'doorjamb', 'towel_rack', 'mouse', 'cupboard', 'door', 'sink', 'bathroom_cabinet', 'home_office', 'soap']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": walk to bathroom, find sink, turn on faucet, put soap in hands, wash and rinse hands,,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <sink>', '[FIND] <sink>', '[FIND] <soap>', '[WASH] <soap>', '[RINSE] <soap>']\n}\n", "input": "\nScene_Objects:['pot', 'beer', 'toilet', 'jelly', 'milk', 'drawing', 'orchid', 'hanger', 'bathroom_counter', 'floor', 'ceiling', 'dvd_player', 'faucet', 'balanceball', 'mousepad', 'towel', 'dining_room', 'wall', 'photoframe', 'light_bulb', 'towel_rack', 'dresser', 'cupboard', 'maindoor', 'sink', 'bathroom_cabinet', 'home_office', 'light', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n", "output": {"thoughts": "I go to the bathroom, open the faucet rinse the hands, wash the hands with soap and rinse again, then wipe with a towel.", "actions": ["[WALK] <bathroom>", "[WALK] <faucet>", "[FIND] <faucet>", "[TURNTO] <faucet>", "[FIND] <hands_both>", "[RINSE] <hands_both>", "[FIND] <soap>", "[GRAB] <soap>", "[WASH] <hands_both>", "[RINSE] <hands_both>", "[TURNTO] <faucet>", "[FIND] <towel>", "[GRAB] <towel>", "[WIPE] <hands_both>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['food_pizza', 'tablelamp', 'food_salt', 'cd_player', 'chair', 'freezer', 'bathroom', 'food_carrot', 'orchid', 'knifeblock', 'floor', 'ceiling', 'phone', 'desk', 'bedroom', 'keyboard', 'remote_control', 'computer', 'window', 'mousepad', 'wall', 'nightstand', 'light_bulb', 'check', 'food_onion', 'curtain', 'mouse', 'door', 'light', 'tooth_paste']\nTask:Listen to music\nresponse:\n{\n    \"thoughts\": Walk to CD player and turn it on then go and sit in a chair and turn the remote control on and then listen to some music.,\n    \"actions\": ['[WALK] <cd_player>', '[SWITCHON] <cd_player>', '[FIND] <chair>', '[SIT] <chair>', '[FIND] <remote_control>', '[TOUCH] <remote_control>']\n}\n", "input": "\nScene_Objects:['glue', 'tablelamp', 'chair', 'walllamp', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'character', 'desk', 'keyboard', 'bookshelf', 'computer', 'ceilinglamp', 'balanceball', 'mousepad', 'pillow', 'wall', 'nightstand', 'bench', 'food_food', 'doorjamb', 'kettle', 'mouse', 'pencil', 'headset', 'light']\nTask:Browse internet\nresponse:\n", "output": {"thoughts": "go to the office, turn computer on, open an browser or any application to open any website.", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['blender', 'chair', 'walllamp', 'detergent', 'milk', 'drawing', 'wallshelf', 'hanger', 'floor', 'ceiling', 'desk', 'bedroom', 'remote_control', 'computer', 'mat', 'tvstand', 'ceilinglamp', 'faucet', 'shower', 'closetdrawer', 'couch', 'pillow', 'wall', 'photoframe', 'television', 'doorjamb', 'diary', 'kitchen_counter', 'dresser', 'napkin']\nTask:Watch TV\nresponse:\n{\n    \"thoughts\": Find the remote, turn on the TV, watch.,\n    \"actions\": ['[WALK] <home_office>', '[WALK] <remote_control>', '[FIND] <remote_control>', '[GRAB] <remote_control>', '[FIND] <television>', '[SWITCHON] <television>', '[TURNTO] <television>', '[WATCH] <television>']\n}\n", "input": "\nScene_Objects:['tablelamp', 'microwave', 'toilet', 'chair', 'trashcan', 'ceilingfan', 'drawing', 'cpuscreen', 'floor', 'ceiling', 'phone', 'desk', 'keyboard', 'bookshelf', 'computer', 'button', 'mail', 'ceilinglamp', 'deck_of_cards', 'balanceball', 'window', 'mousepad', 'wall', 'nightstand', 'photoframe', 'novel', 'doorjamb', 'mouse', 'light', 'dishwasher']\nTask:Write an email\nresponse:\n", "output": {"thoughts": "Switch on computer, Use keybord and mouse to open e-mail provider, type requested e-mail, click send, close e-mail provider, turn off computer", "actions": ["[WALK] <home_office>", "[WALK] <computer>", "[FIND] <computer>", "[SWITCHON] <computer>", "[FIND] <mouse>", "[FIND] <keyboard>", "[FIND] <mail>", "[OPEN] <mail>", "[TYPE] <keyboard>", "[FIND] <button>", "[PUSH] <button>", "[CLOSE] <mail>", "[SWITCHOFF] <computer>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['microwave', 'toilet', 'freezer', 'walllamp', 'bathroom', 'trashcan', 'orchid', 'floor', 'toaster', 'mat', 'faucet', 'shower', 'pillow', 'wall', 'nightstand', 'bench', 'toy', 'coffe_maker', 'television', 'curtain', 'after_shave', 'doorjamb', 'towel_rack', 'kitchen_counter', 'chessboard', 'sink', 'home_office', 'bathroom_cabinet', 'table', 'instrument_violin']\nTask:Go to toilet\nresponse:\n{\n    \"thoughts\": Walk to the bathroom.  Find the toilet.  Lift the top of the toilet.  Sit on the toilet.,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <toilet>', '[FIND] <toilet>', '[OPEN] <toilet>', '[SIT] <toilet>']\n}\n", "input": "\nScene_Objects:['spectacles', 'sheets', 'chair', 'ground_coffee', 'walllamp', 'hanger', 'longboard', 'floor', 'ceiling', 'phone', 'bookshelf', 'keyboard', 'computer', 'stovefan', 'tvstand', 'stereo', 'window', 'clothes_hat', 'wall', 'nightstand', 'bench', 'box', 'food_cheese', 'doorjamb', 'mouse', 'dresser', 'cupboard', 'hair', 'door', 'light']\nTask:Turn on light\nresponse:\n", "output": {"thoughts": "Walk to living room.  Find light switch.  Switch on light switch.", "actions": ["[WALK] <home_office>", "[WALK] <light>", "[FIND] <light>", "[SWITCHON] <light>"]}}
{"instruction": "\nYou serve as an AI task planner. \n1. Your task is to create a plan to achieve a goal by converting it into a sequence of actions. Each action should follow the format \"[verb] <target1> <target2>\", where 'verb' represents the action, and 'target1' and 'target2' are optional arguments.\n2. You must describe your plan in natural language at the beginning. After that, you should list all the actions together. \nEnsure that you response can be parsed by Python json.loads.\n\n\nHere is an example plan to achieve a goal for reference: \nScene_Objects:['pot', 'freezer', 'ceilingfan', 'drawing', 'knifeblock', 'bathroom_counter', 'floor', 'ceiling', 'phone', 'keyboard', 'bookshelf', 'faucet', 'picture', 'window', 'dustpan', 'towel', 'wall', 'nightstand', 'photoframe', 'keys', 'toothbrush_holder', 'food_donut', 'mouse', 'kitchen_counter', 'sink', 'bathroom_cabinet', 'table', 'vase', 'soap', 'hands_both']\nTask:Wash hands\nresponse:\n{\n    \"thoughts\": Cleaning hands,\n    \"actions\": ['[WALK] <bathroom>', '[WALK] <faucet>', '[FIND] <faucet>', '[SWITCHON] <faucet>', '[FIND] <soap>', '[GRAB] <soap>', '[FIND] <hands_both>', '[SCRUB] <hands_both>', '[PUTOBJBACK] <soap>', '[RINSE] <hands_both>', '[SWITCHOFF] <faucet>', '[FIND] <towel>', '[GRAB] <towel>', '[WIPE] <hands_both>', '[PUTOBJBACK] <towel>']\n}\n", "input": "\nScene_Objects:['sheets', 'pot', 'toilet', 'blow_dryer', 'bathroom', 'trashcan', 'orchid', 'dvd_player', 'keyboard', 'mail', 'computer', 'food_bacon', 'stovefan', 'ceilinglamp', 'picture', 'window', 'couch', 'clothes_pants', 'oven', 'photoframe', 'coffe_maker', 'food_food', 'television', 'food_onion', 'coffee_pot', 'sink', 'bathroom_cabinet', 'light', 'dishwasher', 'vase']\nTask:Wash dishes with dishwasher\nresponse:\n", "output": {"thoughts": "grab the vesseles and wash", "actions": ["[WALK] <dining_room>", "[WALK] <bowl>", "[FIND] <bowl>", "[GRAB] <bowl>", "[WALK] <dining_room>", "[WASH] <bowl>", "[WALK] <detergent>", "[FIND] <detergent>", "[RINSE] <detergent>"]}}